I.1-99 - Rigveda Translation: Commentary

Maṇḍala I
[I.1 JPB]
I.2: Vāyu, etc. (Praügaśastra)
The recipients of the various tṛcas making up these two rather simple hymns
are clearly signalled. All three verses making up the first tṛca to Vāyu (vss. 1-3) open
with a voc. vāyo; the second tṛca (to Indra and Vāyu, vss. 4-6) opens with the voc.
índravāyū, while the next two begin with the famous “Vāyav Indraś ca” construction
(on which see Jamison 1988). The pattern is varied in the Mitra and Varuṇa tṛca (79), with the conjoined accusatives mitrám and váruṇaṃ ca opening the first and
second pādas of vs. 7, and the dual dvandva mitrā́váruṇā(v) in second position in the
next two verses, first as a vocative, then as a nominative.
There appears to be some attempt to create bridges between the tṛcas: verses 3
and 4 both sketch a reversal of the usual ritual model; vss. 6 and 7 both concern our
‘insight’ (dhī́).
I.2.2: Ge. suggests that jarante here can be ambiguous, belonging not only to ‘sing’,
but also to ‘awaken’, with identical present stem. This is possible, but only with an
intransitive sense of ‘awaken’: “the singers awaken / wake up to you,” since the
‘awaken’ present is only intrans. (see Gotō 1987: 150). In any case surely the
primary sense is ‘sing’, given the the etymological figure produced by its
grammatical subj. jaritā́raḥ ‘singers’.
I.2.3: The difficult words dhénā and prapṛñcatī́ complicate the interpretation of this
verse. The former, investigated in detail by H.-P. Schmidt (Gd. Nyberg), is now
generally interpreted as ‘(milk)stream’, rendering Geldner’s ‘lip’ and Renou’s
‘tongue’ out of date. As for prapṛñcatī́, the simplex pṛñcatī́ appears in another
Praügaśastra hymn (I.23.16), modifying waters and referring to the mixing of milk
(acc.) with honey (instr.). Given the similarity of context, a direct object referring to
a liquid should be supplied.
What is also puzzling here is in what way the ‘stream’ is Vāyu’s: it should not
originate with him, but rather be destined for him, but then why is the ‘pious man’
apparently receiving the benefit of it? As in the next verse, there seems to be a bit of
role reversal here, with the gods depicted as providing the ritual benefits rather than
receiving them. Presumably the point is that the pious man gets the benefit
secondarily, by having pleased the god, but the dative dāśúṣe is striking, esp. as it is
apparently parallel grammatically (though it cannot be functionally) to the dat.
sómapītaye ending the verse. Although the verse presumably depicts the sacrificer’s
offering of soma to Vāyu to drink, the lexicon and the case usage complicate the
message.
I.2.4: As noted in the comment on the last verse, the ritual model here is shaken up a
bit: Indra and Vāyu are urged to come with práyobhiḥ, a word generally used of
‘pleasurable offerings’ that are presented to the gods and to which they come (cf.
VIII.60.4 abhí práyāṃsi .. gahi). Our translation “with delight,” agreeing with most
other translators, avoids, and conceals, the problem. The gods should not be bringing
práyāṃsi. The little disturbance of the ritual model is confined to these two verses in
this hymn.
The pāda-final position of hí here is unusual, and I have no explanation for it,
esp. as it does not take second position in its clause as is usual.
I.2.7-9: As mentioned in the intro., this tṛca contains the trio dhī́ ‘insight’ (vs. 7),
krátu- ‘intention’ (vs. 8), and dákṣa- ‘skill’ (vs. 9), the three elements necessary to
conceive and carry out an action. Their interconnection is emphasized by the fact that
all three are in the accusative and each is stationed initial in the last pāda of its verse.
The juxtaposition across vss. 8-9 of krátum (beginning 8c) and kavī́
(beginning 9a) may also be meant to evoke the well-established compound kavíkratu- ‘having the will/resolve of a poet’, ‘having a poet’s purpose’, an occurrence of
which is found in the preceding hymn by the same poet (I.1.5).
I.2.7: Here and everywhere else it is found, the word riśā́das-, an epithet of various
gods, is opaque. There are currently two competing and entirely different
interpretations: that of Karl Hoffmann (Aufs. 564 n. 16) as ‘discriminating, fastidious’
(< ‘picking at food’) and Paul Thieme’s ‘caring for the stranger’ (Fremdling). See
EWA s.v. The contexts are not diagnostic, and it is probably the case that the epithet
was no longer understood even as it was being deployed (note that it is almost always
pāda-final, possibly a sign of formulaic freezing).
Throughout our translation we have followed the Thieme interpretation, but
not with any great conviction. One thing in favor of the Thieme interpretation is that
the word is regularly applied to one or more of the Ādityas (as here), who might be
expected to show care for humans in their charge. That it is also regularly used of the
less ethically inclined Maruts might give us pause (though these contexts are
generally benevolent ones) – except that ‘fastidious’ is even less a likely quality of
the Maruts than ‘caring for the stranger’.
I.3 (Praügaśastra continued)
As in I.2 the recipients of the various tṛcas are emphatically signalled. In vss.
1-3 to the Aśvins, the voc. áśvinā opens the first two verses, while their alternative
name nāsatyā opens the second pāda of the third. The voc. índra opens all three
verses of the next tṛca (4-6). The Viśvedevāḥ tṛca contains three instances of that
phrase: the voc. in 7b, nominatives opening vss. 8 and 9. The final tṛca to Sarasvatī
likewise contains three occurrences of her name in the nominative, but all three end
their pādas (10a, 11c, 12a).
I.3.2: śávīra- rendered as ‘powerful’ in the publ. tr. But see disc. below ad I.30.17.
dhíṣṇya- and related forms are obscure and much discussed; indeed Ge.
refuses to translate the word. We generally follow the view of Pinault (UTexas Vedic
2
Workshop), who takes it to mean ‘related to / proper to the holy place’, thence
simply ‘holy’.
I.3.3: In the compound rudra-vartanī, number is of course neutralized in the first
member. The Maruts are regularly called Rudras (without vṛddhi or derivational
suffix) after their father. The ‘course of the Rudra/Maruts’ is simply a reference to
the midspace (antarikṣa) much frequented by the Maruts, where the Aśvins are now
driving.
I.3.8: A small grammatical mismatch here: the phrase viśve devā́saḥ and the
adjectives modifying it (aptúraḥ, tū́rṇayaḥ) are nominatives and should not be the
subject of the imperative ā́ ganta. Ge. (and WG) ignore the problem by translating
the nom. as voc. (“Ihr Allgötter”). Although the effect is minor, my translation
reflects the grammatical disjunction by rendering pāda b as an interjection.
Another question is why 7b contains the same 2nd pl. imperative, except with a
different grade of the root: ā́ gata vs. ā́ ganta. Both forms are reasonably well
attested, with 7b a repeated pāda (=II.41.13a, VI.52.7a). Whatever the history of the
distinction, the synchronic distribution seems to be metrical, with ā́ gata almost
always final, providing an iambic cadence in dimeter verse, and ā́ ganta found earlier
in the verse.
In b tū́rṇayaḥ was carelessly omitted from the tr., which should read “Come
here swiftly…”
I.3.9: I follow the analysis of the hapax éhimāyāsaḥ as a frozen 2nd sg. imperative
phrase, “éhi mā́+yāḥ” (“come! don’t go”), transformed into an adjective in the nom.
pl. masc. – an analysis that goes back at least to Sāyaṇa. Ge also follows this analysis,
though it is somewhat difficult to excavate from his “willkommen und ungern
fortgelassen.” I interpret it as representing the words of the singers’ invitation
regularly heard by the VDs. The other currently competing explanations, as a frozen
phrase “éhi māyā” [better voc. māye?] “come here, magic” (Old) or as a deformation
of áhi-māya- ‘vielgestaltig’ (Gr) [=‘snake-sly’ (J+B)] (BR, followed by Gr), fit less
well into the content of the hymn, which after all focuses on calling the various gods
to the ritual; note the ā́ gata, ā́ ganta of vss. 7-8 addressed to the same VDs. Support
for this analysis may also come from the next hymn (I.4), attributed to the same poet,
in which successive vss. (3c, 4a) contain the imperatives ā́ gahi ‘come here’ and
párehi ‘go away’, with at least the former addressed to the god Indra.
I.3.11-12: Note the contrastive values of the simplex pres. cétantī ‘perceiving, taking
note’ and the -áya-pres. (prá) cetayati ‘makes perceived, reveals’ in successive vss.
I.4 Indra
3
I.4.2: godā́(ḥ) of pāda c echoes godúhe of 1b. I consider pāda c a proverbial
expression – when a rich man is pleased, he gives cows – though it’s obviously
applied to Indra here.
I.4.4: Striking is the abrupt change of subject of the 2nd sg. imperatives, from Indra
(3c) to an unidentified human companion (4a).
My interpretation differs substantially from those of most others in pāda b. In
my view, the accusative índram marks Indra as the one directly interrogated, rather
than (with most interpretations) the one to be asked about. Most interpretors take
vipaścítam as identifying the person to be interrogated (e.g., Ge. “einen Weisen”),
thus assuming two different referents for the accusative singulars in that pāda: “ask
the wise one about Indra.” I find that unlikely, in part because, though vipaścít- can
be used of humans, it more often qualifies gods.
Taking Indra as the one interrogated has further effects on the interpretation.
For others the relative clause in c has Indra as its subject (yáḥ) and the 2nd ps. te
refers to the human interrogator: it is Indra who is dear(er) to you, the poet, than your
comrades. I, on the other hand, take pāda c as a syntactic hybrid, with an underlying
direct discourse question, directed to Indra, “who [expected káḥ] is your choice from
among your comrades?” incompletely converted into a relative clause in indirect
discourse “ask (Indra) about (the one) who [yáḥ] is your [=Indra’s] choice…” In my
view the 2nd ps. ‘your’ of “your choice” in pāda c refers to Indra, not to the subject of
the imperatives párehi and pṛchā of ab, while Indra is in the 3rd ps in pāda b. (I will
not even contemplate the possibility that pṛchā in b is a 1st ps subjunctive: “Go away.
I will ask / let me ask Indra…”)
Although this interpretation complicates the syntax, in my opinion it fits
better into its tṛca and better reflects the relationship between Indra and humans. As
often in Indra contexts, the poet worries that Indra will favor others over the poet
himself, and this verse poses the question directly to Indra: who do you like best?
Indra’s presumed and desired answer is “you!” This answer then allows the poet to
dismiss those who criticize him for not spreading his devotion around to other gods
(vs. 5) and defends this exclusive focus as a good bargain, as the rest of the world has
to admit (vs. 6).
I should admit, however, that the standard view is somewhat compatible with
my larger interpretation, in that “… ask about Indra, who is your [=poet’s] choice
from among your comrades” could reinforce that message that our focus is only on
Indra, not on other gods. But I do not see how questions about Indra fit with the next
two verses.
I.4.5-6: Most interpretations take these two verses as syntactically parallel (e.g., Ge.
“Mögen … Und mögen …”), but the impv. bruvantu and the opt. vocéyuḥ are surely
doing different things: the imperative is concessive: “let them say / even if they
say ...” while the optative expresses the conclusion that the rest of the world would
have to draw. The parallel utá’s that open these verses might give us pause, but they
may have something like the value “on the one hand … on the other.”
4
I.4.7: (JL) The b and c pādas both end with an adjective modifying the implicit object
soma, a compound of the root √mad ‘exhilarate’ (the second time in its byform
mand) and a noun expressing the personal object of the verb, but in exactly opposite
order: nṛ-mā́dana- and mandayát-sakha-, what might be called a “compound
chiasmus.” A less complex etymological figure is found at the beginning of the
verse: āśúm āśáve.
I.4.8: My occasional tr. of ghaná- as ‘bane’ was inspired by my husband’s treatment
of etymologically related nominal constructions in Greek and Germanic (Watkins
1996: 418ff., 423). I think JL for reminding me of this.
I.4.9: (JL) Etymological figure also in vā́jeṣu vājínam, immediately followed by
vājáyāmasi, which, however, is synchonically distinct from the ‘prize’ words.
I.4.10: There may be bit of ring composition here, with 10b supāráḥ echoing the first
word of the hymn, 1a surūpa-.
I.5 Indra
I.5.1: Seems deliberately to echo last vs. of preceding hymn (I.4.10), with pāda b
índram abhí prá gāyata “sing forth to Indra” matching I.4.10c tásmai índrāya gāyata
“sing to him, to Indra” (the difference in case being governed by the presence of the
preverb abhí in I.5.1). I.4.10 is then exactly repeated in I.5.4c. The sákhāyaḥ of I.5.1c
also recalls I.4.10b sákhā — though the latter refers to Indra and the former to the
priest-poets. But I.4.4c contains a pl. sákhibhyaḥ, which in our analysis has the same
human referents as I.5.1, showing the reciprocal relation between men and gods that
was one of the points of I.4.
I.5.2: purūtámam purūṇā́m is pleonastic, meaning literally “the first of many, of the
many ones.”
I.5.3: My interpretation of these sentences as questions is not overtly marked in the
text, but seems a reasonable use of the subjunctives.
I.5.5: The double dative sutapā̎vne … vītáye with yanti is more literally “… go to the
soma-drinker [lit. ‘pressed (soma) drinker’] to pursue (him).”
I.5.10: A whiff of ring composition – 10c ī́śānaḥ ‘having control over’ echoes 2b
ī́śānam, both modifying Indra. In 2b the god controls something undeniably positive,
“choice things,” which he will presumably distribute to this favorites. In 10a he
controls “the deadly weapon” that other mortals might wield against us. The identity
of expression ties together the very different sentiments.
5
I.6
As noted in the intro. the Anukramaṇī’s identification of the divinities as
Indra (1-3, 10), Maruts (4, 6, 8-9), and Indra and Maruts (5, 7) does not conform to
the content of the hymn, which is quite disjointed, but appears to concern, at least in
part, the Vala myth. The Maruts do not seem to figure at all in the hymn; the plural
entities with Indra are probably the Aṅgirases. For my view of the structure (which is
informed by the discussions of Ge and Old), see my intro.
I.6.1-2: These verses begin identically (yuñjánti ‘they yoke’), inviting the audience to
equate the action of the two verses.
1b: The referent of the apparent acc. plural tasthúṣaḥ ‘(those) standing still’ is
not given. Ge. (/WG) thinks it refers to stable things on the earth, but if the sun is
referred to in the first pāda, it is more likely to “move around” celestial features than
earthly ones, and the stars or other luminous heavenly bodies are referred to in the
next pāda. Re’s interpretation (flg. Ludwig) of tathúṣaḥ as an abl. sg. (“from the one
standing still” – “à partir de (l’espace) immobile”) is ingenious and would match the
minor syntactic idiom ‘yoke from ABL’ (e.g., I.115.4 yadéd áyukta harítaḥ
sadhásthāt), so it cannot be dismissed. The ablatives in the final verses (9-10) might
lend weak support for Re’s view.
2b: The hapax compound vípakṣasā is difficult and has been variously
interpreted. The second member, pákṣas- (and related and more common pakṣá-),
can mean either ‘wing’ or ‘side’; the first member, ví-, is most likely the preverb ví,
but in compounds this element has a number of possible meanings: ‘without’, distant’,
‘wide’,‘alternating/opposite/different’, ‘dispersed’. It could also possibly represent
ví- ‘bird’, which has been claimed as the first member of some other compounds (see
EWA s.v. váy-, KEWA III.266). The possible combinations of these two ambiguous
elements allow for a number of interpretations. I more or less follow the Sāy/Gr
interpretation, ‘auf beiden Seiten des Wagens gehend’, though I take it as an
adverbial instrumental, not a dual. (The presence of a number of duals in -ā in the
verse does not favor an adverbial interpretation, however.) Re’s “aux ailes d’oiseau”
obviously takes the first member as the ‘bird’ word, while WG “die mit weiten
Flügeln” takes ví as the preverb, but with the second member meaning ‘wing’ as in
Re’s interpretation. Ge’s ‘auseinanderstrebenden(?)’ treats the second member quite
loosely.
I.6.3: The baffling part of this verse is the voc. plural maryāḥ ‘o young men’ in b,
embedded in a verse that otherwise has 2nd singular reference (ajāyathāḥ c, along
with sg. ptcpl. kṛṇván in a). There is no clear referent for this voc., though it may
refer to the unidentified plural subjects of the verbs in vss. 1-2 (yuñjánti) and 4
(eriré). In the plural márya- is often used of the Maruts, which may account for the
Anukramaṇī identification of them as divinities of the hymn. Though Ge suggests
they may constitute the audience for the singer and Re that they are the singers
themselves, this seems unlikely because when márya- has an identifiable referent, it
6
is never a human. I tentatively assume that it refers to the Aṅgirases as the firepriests who first kindled Agni, the subject of the verse.
As for the subject of ajāyathāḥ, contrary to most interpretations I take this as
primarily referring to Agni, not the sun, though perhaps, with Re, “Agni solaire.”
I.6.4: Grammatically problematic is the accent on the verb eriré in b, a fact that
seems to have been elided in most translations, including mine. Old suggests that the
particle áha may have conditioned the accent, but this seems unlikely because áha
doesn’t have this effect elsewhere. I would now alter the translation to make vs. 4
syntactically dependent on vs. 3, without an overt subordinator. Thus, “you were
born together with the dawns, / (as/when) just after that they once again roused …”
Again, though most commentators (save Old) consider this to concern the rebirth of
the sun, I think it more likely that Agni/the ritual fire is the object, esp. as erire +/- ní
is regularly used of establishing the ritual fire (e.g., I.134.4). As for c, the service to
the ritual fire of the unnamed subjects (=Aṅgirases?) would account for their
receiving a name worthy of the sacrifice; see, e.g., I.72.3.
I.6.5: As noted in the intro., this verse helps resolve the unclear referents in the
earlier part of the hymn by giving a relatively clear sketch of the Vala myth, with
Indra finding the cows after his companions “break the stronghold (=Vala).”
I.6.6: This verse contains, in my opinion, what Re might call a “legère zeugma,” in
which the verb anūṣata “they bellowed” takes (as is usual) an acc. of the target of the
bellowing (“to the finder of goods,” i.e., Indra) in the frame, but in the simile it takes
an acc. of the content of the bellowing (“their thought”). Ge avoids this mismatch of
acc. function by removing matím from the simile by supplying a form of √bhṛ ‘bear,
present’, leaving devayántaḥ as the only term directly compared in the simile: “Wie
Gottverlangende, die das Lied [vortragen], so haben die Lobreden … hergerufen.” It
is certainly true that matí- is common as the object of √bhṛ and that √nu doesn’t
normally take an acc. of content, but since the poet of the hymn has pushed the
linguistic limits elsewhere, I prefer to think he meant the jarring figure. Note that
there is also a mismatch between the two subjects, with the simile referring,
implicitly, to human actors, while the frame has ‘songs’ (gíraḥ) as subject (unless we
take the Aṅgirases or the cows as subj. and allow anūṣata to take two accusatives:
“they bellowed their songs to the finder of goods” – however, √nu doesn’t take two
acc., to my knowledge).
The zeugma may iconically represent the fact that the verse connects across a
temporal gap as well: the simile seems to refer to present-day worshippers producing
their praise, but the frame (with augmented verb form) refers to the mythic past of
the Vala tale. This verse thus serves as a transition to the here-and-now of the current
ritual, which is treated in vs. 7.
7
I.6.7: As noted in the intro., this verse pairs structurally with vs. 3; I therefore take
Agni to be the subject, with the verse expressing the kindling of the fire at the time
when Indra arrives to receive the morning offering.
The form dṛ́kṣase is isolated, but its grammatical identity is fairly clear (see
Narten, Sig. Aor. p. 146): a 2nd sg. mid. s-aor. subj. with the “wrong” grade of the
root (expect *darkṣase); it is probably based immediately on the other s-aor. middle
form, 3rd pl. indic. adṛkṣata (5x, once accented) of the same metrical shape (minus
augment), which always appears final, as does dŕkṣase, and usually in dimeter verse
as here.
The two beings in pādas ab, one as unexpressed sg. subj. of the verb, one in
the instr., are then referred to as a pair in the du. nom. of pāda c, the predicate of an
unexpressed nominal sentence “you two are…”
I.6.8: The Aṅgirases are presumably the referents of the instr. phrases, and the verse
is, like 5, a pretty clear allusion to the Vala myth.
I.6.9: As suggested in the intro., this is the last real verse of the hymn, as vs. 10 is a
mere variant of 9, and it shows a bit of ring composition: the diváḥ … rocanā́t echoes
rocanā́ diví of 1c, and if we were to accept Re’s interpr. of 1b tasthúṣaḥ as an
ablative, the ablatives átaḥ and rocanā́t would match it grammatically.
This is a rare example of the present middle ṛñj not taking an acc. (see Tucker
2002: 284 n.17, HS 115 “RV rgmín-, ṛgmíya- and ṛñjate”). (JL)
I.7 Indra
I.7.1: anūṣata provides a link to the immediately preceding hymn, I.6.6.
I.7.2: Though a number of interpreters (Gr, WG, Scar) take vacoyújā as an instr. sg.
and supply ráthena ‘chariot’, this form otherwise (4x) is only du. and modifies hárī
‘the two fallow bays’. Thus, it seems better to follow the Sāy/Ge/Re interpr. As Ge
points out, the untethered ā́ in b allows a form of √sthā to be supplied, in the idiom ā́
√sthā ‘mount’. The verse is then slightly unusual in referring to Indra’s twin horses
in two grammatical cases in the same sentence (háryoḥ loc., [hárī] vacoyújā acc.).
I.7.3: A more felicitous tr. of dīrghā́ya cákṣase might be “to be seen for a long time,”
but “for the long view” allows the phrase to be read as referring to either time or
space (“to be seen for a long distance”) or both.
The usage of the instr. góbhiḥ is somewhat strange; it is clearly not meant
either as an instr. of agent/instrument or of accompaniment, at least of simple
accompaniment. It might be an instr. of separation, or, as in this tr., an adjunct or
accompaniment to the obj.: “the rock (which was) with cows.” ET points out to me
that such a construction would be very unusual; I suggest that it could derive from an
instr. of accompaniment: “the rock along with its cows.”
8
I.7.6: For the pot, see intro. The doubling of the 1st pl. pronoun (naḥ in a, asmábhyam
in c) is probably simple redundancy, with naḥ a Wackernagel placeholder at the
beginning of the sentence, anticipating the full pronoun that opens c. However the
naḥ could possibly be construed with the voc. sátrādāvan ‘who give in every way’ in
b, though it seems a bit distant from the enclitic.
I.7.7: Improper relative, as shown best by Re’s rendering, “Les corps-de-louange qui,
poussée, (vont toujours) plus haut…” (Re’s suspension dots). The masc. nom. pl. yé
… stómāḥ of ab has no matching grammatical referent in the main clause of c,
though it is picked up by its semantic and etymological equivalent, fem. sg. suṣṭutí-.
I.7.8: Connected to vs. 6 by shared vocab., vṛ́ṣan- (a) and (the rather rare)
ápratiṣkuta-, though separated by vs. 7.
I.7.9: Incomplete sentence, consisting only of rel. cl., completed by main cl. of 10.
The ékaḥ opening this last sentence of the hymn and the kévalaḥ ‘exclusively’ that is
its last word are more insistent counterparts of íd in the opening pādas of vss. 1-2.
Once again Madhuchandas seems to be faintly signalling ring composition.
I.8 Indra
I.8.2: Incomplete sentence, with relative hanging off rayím ‘wealth’ in the previous
verse. Two methods of fighting are contrasted: ‘fighting (-hatyā́-) by fist’ (muṣṭi-, my
‘bare-knuckled’) and ‘on horseback’ (árvatā). Although the two terms are
grammatically parallel (instrs. muṣṭi-hatyáyā and árvatā), they are not semantically,
since it’s the first member of the compound, muṣṭi- that corresponds to árvatā, and
‘fighting’ must be supplied with the second term.
The verse shows overt signs of late grammatical features: esp. the –ai ending
of the middle subjunctive ruṇádhāmahai (rather than –e), but also the longer –ā-stem
instr. –áyā (rather than –ā́).
I.8.3: Concatanation of tvótāsaḥ (pāda a) with the same form in 2c, though the one in
2c requires distraction (tuvo-), but not the one in 3a. Ge/Re take ghanā́ as (an
archaic) instr. sg., but nom. pl. ghanā́(ḥ) seems preferable, esp. as Madhuchandas
uses the same word in the sg. as a personal designation in I.4.8 (where it applies to
Indra and which I tr. ‘bane’). So Old. Although designating animate beings (namely
“us”) as “hammers” may seem unusual, it’s not unprecedented, at least in English: cf.
the rock song entitled “Sometimes you’re the hammer and sometimes you’re the nail,”
and (gleaned from Google) a quotation from an American poet unknown to me,
Edwin Markham (1852–1940), “For all your days be prepared, and meet them ever
alike. When you are the anvil, bear - when you are the hammer, strike.”
I.8.6: The whole verse is a relative clause with accented verb (yá ā́śata), with no
overt antecedent available in either the preceding or the following vs. My solution
9
follows Old, who suggests that it implicitly hangs off vs. 5: Indra’s power is (for
those) who… This fits the message of the hymn, that men’s success is entirely
dependent on Indra’s aid and intervention, a message that is reinforced by the
interdependence of various vss. already noted (1-2 [main cl., rel. cl.], 2-3 [lexical
concatenation]) and to be described below [7-10].
Though ā́śata lacks expressed obj., ‘him’ (=Indra) should be supplied, on the
basis of passages like I.85.7, VIII.97.9.
I.8.7: Yet another untethered rel. cl. In my view, the description of Indra’s physical
capacity serves as the basis for the expressions of Indra’s vast liberality and help in
vss. 8-9, each of which begins with evá hí “for just in the same way.” Therefore the
yáḥ ‘which’ of 7a seems a substitute for yáthā ‘even as’, the usual relative with evá.
Although I do not so translate it, 7-8 could be rendered “even as his cheek … swells
…, even so is his liberality…”
For kukṣí- as ‘cheek’, not ‘belly’, see Jamison 1987 (Gs. Cowgill).
I.8.8: The image of Indra’s generosity as “a ripe branch” is an unusual one; I do not
know of a parallel. In any case, “ripe branch” must be a condensed expression for
something like a branch laden with ripe fruit. (JL)
I.8.10: Despite sharing the evā́ hí opening with vss. 8-9, this verse is not entirely
parallel with those two, which express the vastness of Indra’s liberality and help.
Here it is what we owe Indra, praise and recitation, that are implicitly suggested to be
as vast as what he gives us. A tr. more parallel to the previous two verses would be
“Just the same [that is, just as vast] are those things beloved of him, the praise-song
and recitation to be proclaimed …” However, I favor the published tr., with śáṃsyā
‘to be proclaimed’ as predicate, because it provides a hortatory end to the hymn.
I.9 Indra
I.9.1: somapárvan- ‘soma-joint’ could refer either to the segments of the stalk of the
soma plant (e.g., Re) or to the segments of the Soma Sacrifice (e.g., WG): Ge
suggests it’s a word play. It is difficult to judge, but I weakly favor the horticultural
interpretation.
There is no explicit 2nd ps. in c, but the general interpretation of this pāda as
referring to Indra seems correct.
I.9.2: For the doubling of the enclitics īm enam see Jamison 2002 (pp.).
I.9.6: With Re I take the two acc. pl. –vant-adjectives (rábhasvataḥ … yáśasvataḥ) as
proleptic, with the acquisition of these qualities being the result of Indra’s impelling
of us – rather than taking them as qualities we already possess, as most translators do.
10
I.9.8: rathín- should of course mean ‘possessing chariots’ or express some looser
association with a chariot or chariots (such as Re’s “carried on chariots”) but since
there’s no obvious association of refreshments with chariots, an idiomatic and
figurative use like Ge’s “wagenvoll” seems appropriate – hence my “by the cartload.”
I.9.9: In my view gṛṇánta(ḥ) is an instance of the comparatively rare (but more
common than generally supposed) predicated present participle. Other translators
(Ge, Re, WG) take the participle as attributive and consider the sentence incomplete.
I.9.10: As Thieme (Fremdling, pp. 11f.) points out, the verse sets up an implicit
contrast between Indra, who is “at home” (nyòkas-) wherever soma is pressed, and
the stranger – but this opposition also implicitly suggests that, despite being a
stranger or foreigner, any man can offer soma and praise to Indra, who will make
himself at home in those foreign parts. This contrast would be better expressed by
“even the stranger chants…” rather than the published “the stranger himself chants…”
The position and function of ā́ (embedded in éd) in b are unclear. The verb
√arc doesn’t take ā́ and in any case preverbs don’t usually ended up stranded in the
middle of a pāda (of course the etymological figure bṛhád bṛhaté could have been
fronted around it); a mid-pāda position suggests a role as adposition, but as an
adposition ā́ doesn’t take a dative.
I.10 Indra
I.10.1: The first three pādas almost, but not quite, provide a tripartite ritual speech
division: Sāmaveda, Ṛgveda, X? Veda. The last is the problem: the “formulators”
don’t work very well as speakers of Yajurveda yajuses, and it’s too early for the
brāhmaṇa priest to be associated with the Atharvaveda, as in later Vedic.
Pace most translators, pf. yemire is ordinarily presential in value; see Kümmel
s.v.
I.10.2: Most translators take the subject of ab to be the sacrificer, but Indra seems a
more likely candidate, esp. since kártva- ‘to be done’ is regularly used of the
prospective deeds of Indra (e.g., II.30.10, IV.18.2, VIII.63.6).
I.10.3: On hí with the imperative marking that clause as the causal basis of the next
clause, here initiated by the logical connector átha, see Brereton 2012 [Bronkhorst
Fs.].
I.10.5: rāráṇat: pf. subj. with presential value, like the whole pf. system of this root.
See Kümmel s.v. and Jamison (JLGR Fs).
I.10.6: The case usage here is somewhat odd, in that the three benefits we beg Indra
for, in strict parallel structure, are in loc., dat., and loc. respectively (underlying
11
forms sakhitvé … rāyé … suvīrye). However, all end in -e — showing that surface
phonetic agreement can sometimes trump case function.
The transformation of an epithet (śakrá ‘able’) into its associated verb (śakat
‘he will be able’) is a neat little figure and demonstrates the importance of gods’
dynamically living up to their verbal attributes. (For the almost identical pāda see
VIII.32.12.) It is an example of a type of verbal transformation of divine epithets into
desired divine action that Elizarenkova (1968: 267–68) attempted to claim as the or
an organizing principle of RVic hymnic composition. Of the other standard
translations, only Re (“… le puissant; qu’il exerce .. sa puissance”) captures the
etymological figure.
I.10.7: The two compounds in pāda a do not occur elsewhere and are grammatically
and interpretively ambiguous. The semantic reference of the two words is clear —
the easy opening (ví √vṛ) of the Vala cave and the easy driving out (nir √aj) of the
cows, using the standard lexemes for those actions — but 1) what are their stems?
and 2) assuming they are adjectival, what do they modify? Gr/Lub analyse them as a-stems -- also AiG II.1, though AiG II.2 takes sunirájam as belonging to an a-stem
(p. 86) but suvivṛ́tam to a root noun (p. 43) [and Hauschild’s Index to AiG lists them
both as root nouns, somewhat emphatically] -- while, e.g., Old and Scar take both as
root nouns. There is another formal anomaly: the pāda they form, suvivṛ́taṃ
sunirájam, has only one internal heavy syllable, the final syllable of the first word,
where the initial consonant of the second makes position. A very unusual metrical
line. Arnold (VM 125–26, 290) suggests the possibility of reading suvīvṛ́tam on the
basis of the lengthening of the final vowels of the preverbs ápi, abhí, pári, etc.,
before forms of √vṛ. Thanks to ET for pointing this out.
As to their reference, the general approach has been to take them as
modifying an unexpressed índram, supplying the whole structure of 6a (tám …
īmahe “we beseech him”) or some similar verb phrase to provide a grammatically
acceptable referent for the two forms in 7a. But this solution is not very satisfying:
Indra appears in the 3rd ps. nominative in 6c and as 2nd ps. vocative and subject of
impv. in 7bcd, so extracting an acc. from a pāda in the past seems arbitrary. Scar
suggests that the two words might instead modify yáśaḥ in b, which has the merit of
providing a referent close by; however, this would technically eliminate the
possibility that the two are root nouns, since yáśaḥ is neut. and presumably nom., and
if the two words in pāda a are root nouns, they can only be acc. sg. I am nonetheless
attracted by this solution (and would therefore be open to the -a-stem interpretation),
with the possible modification that the two might actually be nouns (“the easy
opening …, the easy driving …”) that specify the glory (yáśaḥ) of b. I have not
troubled myself to figure out how the accent and other details of the formation would
work, however.
On the possible double sense of tvā́dātam in b, see intro.
I.10.8: jéṣaḥ: The standard translation take this s-aor. subj. as a functional impv.
parallel to dhūnuhi in d, whereas I take it as having real subjunctive value. This has
12
the merit of providing a main clause to the subordinated hí clause of ab. Moreover,
the otherwise identical pāda VIII.40.10 with 3rd sg. subj. jéṣat has clear subjunctive
value, and in addition there is already a well-attested “-si imperative” jéṣi (7x) that
fills that function for the s-aor., so it seems unlikely that jéṣaḥ would be so used.
I.10.9: ā́śrutkarṇa śrudhī́… shows the same transformation of an epithet into a
derivationally related divine action as 6c.
I.10.10: Takes the verb phrase of 9a śrudhī́ hávam and elaborates on both its
members, with b havanaśrútam and c hūmahe.
I.10.11: Kuśika is the ancestor of the Viśvāmitras, the family to which our poet
belongs. As this is the last hymn attributed to Madhuchandas in this set, an ancestral
reference is in order.
I.11 Indra
I.11.1: The phraseology involving “songs” (gíraḥ) and “strengthening” (√vṛdh)
matches that of the last verse of the preceding hymn (I.10.12) attributed to the father
(or other ancestor) of this poet.
“Lord of prizes and lord of settlements” in d may set up an implicit between
battle and battle-like activity (contests with prizes) and peace.
I.11.2: “Conquerer” (jétar-) as epithet of Indra here may be responsible for the poet’s
name Jetar in the Anukramaṇī.
I.11.3: Though the printed text reads yádī ‘if’, with (as often) lengthened final vowel,
nothing prevents us from taking this as yád ī, ‘when’ + enclitic acc. pronoun,
anticipating the expressed acc. obj. For this phenomenon, see Jamison 2002.
I.11.5: Though it may seem odd that the enemy Vala is called ‘fearless’, the other
solution, to take ábibhyuṣaḥ as an “irregular” nom. pl. masc. pf. part. (expect
ábibhīvāṃsaḥ) (so Sāy/Old) is not satisfying, esp. as it’s hard for me to see how the
gods could be both “fearless” and “being pushed back” (although ET points out that
they could have come to Indra’s aid without fear, even though being pushed back).
The question is what syntactic function the gen./abl. ábibhyuṣaḥ is fulfilling. Re
seems to take it as a gen. agent with tujyámānāsaḥ (“pressés par le (dieu) sans peur”),
but gen. agents are rare and confined to semantically and grammatically specialized
situations (see Jamison 1979), and an ablative value “before,” as implied by Ge,
seems better. Since expressions of fearing take the ablative, we can even assume an
underlying implicit contrast: “the gods, (fearful) of [=from] the fearless one…”
I.11.6: Although Ge suggests simply that the poet is announcing Indra’s gifts to the
river of his native land, this may have a further mythological reference. Esp. in
13
X.108 (Saramā and the Paṇis), the (Vala) cave in which the Paṇis have trapped
Indra’s cows is on the edge of the world, across the river (Rasā) that borders the
world. Here the poet may be evoking this myth to indicate the efforts that he (and
Indra) must expend to retrieve the good things his community desires, and to
emphasize that poets and wise men (see also vs. 7) must bear witness to Indra’s
deeds performed far away in order to attract his munificence.
I.12 Agni
I.12.6: The plethora of ritual fires implied by the āmreḍita agním-agnim in vs. 2 is
made more explicit in this expression of the kindling of one fire by another,
presumably (as Ge suggests) through the taking out of the Āhavanīya fire from the
Gārhapatya, much treated in the later ritual lit. The Āhavanīya may be referred to in
5a ghṛ́ta-āhavana- ‘whose oblation is ghee’, and in this vs. the second pāda (6b)
might contain allusions to the three ritual fires, gṛhápati- ‘houselord’ a transparent
reference to the Gārhapatya and yúvan- ‘youth’ referring to the newly kindled
Āhavanīya (see pāda a). However, this would leave kaví- ‘poet, sage-poet’ as a
designation of the Dakṣināgni, which doesn’t make a lot of sense, as far as I can see.
I.13 Āprī
I.13.5: the singular amṛ́tasya ‘of the immortal’ seems to refer to the collectivity of
gods who will come to the sacrifice and sit on the barhis. Vss. 7, 9 name some of the
individual gods who will sit on the barhis.
I.14 All Gods
I.14.3: All these gods names are in the accusative, but there is no verb to govern
them, either in the verse or in the immediately preceding or following pādas. One
solution is to reach back to 2a ā́ ACC káṇvā ahūṣata “The Kaṇvas have called ACC
here,” though skipping over the intervening syntactic constructions is not appealing.
However, the recurrence of a similar construction in 5ab īḷate ACC … káṇvāsaḥ “The
Kaṇvas solemnly invoke ACC” may suggest that the structure of invocation underlies
the hymn.
I.14.6: By making pāda a a nominal sentence, from which the relative clause of bc
hangs, I avoid the need to supply main clause verb for c that other tr. encounter. E.g.,
Ge “Die … Fahrrosse, die dich fahren, (mögen) die Götter … her(fahren).”
However, in the following vs. (7) Agni himself is urged to bring the gods here, so
lumping together the transport of Agni and the gods as in my interpretation of 6 may
not be in the spirit of their separation in 7. However, I still feel that the syntactic
argument is strong.
14
I.14.7: The ab pādas literally mean “make (the gods) possess wives,” but we know
from the ritual that this refers to the coming of the gods along with their wives. Cf.
III.6.9 pátnīvataḥ … devā́n … ā́ vaha. Ge translates our phrase literally: “Die
Opferwürdigen … mache beweibt,” but then paraphrases it in his note: “D. h. bring
ihre Frauen mit.” The problem would be solved by supplying the preverb ā́, because
ā́ √kṛ generally means ‘bring/attract here’. Although I am generally loathe to supply
material without a clear warrant, it is the case that the immediately preceding pāda,
6c, begins with ā́, which might have been taken to have domain over what follows.
I.14.9: The hapax ā́kīm (so, e.g., Gr, Aufr, HvN, Lub) or, more likely, ā́ kīm contains
what is apparently a now meaningless particle kīm, matching the shape of the acc.
particles īm and sīm. Though clearly derived from the interrogative *ku̯-stem, it has
lost all interrogative value, presumably “laundered” through the weak negative
indefinites ná kīm (or nákīm) and mā́ kīm (or mā́kīm). It is not at all clear what, if
anything, kīm is doing here.
I.14.10: Instrumental plurals begin (víśvebhiḥ) and end (dhā́mabhiḥ) the verse. The
question is whether they should be construed separately or together. Ge chooses the
latter path: “Mit allen Verkörperungen des Mitra (der Freundschaft).” But because
the hymn is dedicated to the Viśve Devas and there is an emphasis on them
throughout (see esp. vss. 1, 9, with ‘gods’ or words referring to them collectively in
vss. 2, 6, 7, 8), I prefer to supply ‘gods’ with víśvebhiḥ and take dhā́mabhiḥ
separately (sim. Re, WG), interpreting mitrásya not as the god’s name (or not
principally the god’s name) but as referring to the alliance that undergirds the
sacrificial system.
I.14.12: The construction involving the normally causal particle hí and the imperative
is a troubling one. Brereton (2012 Bronkhorst Fs.) plausibly argues that in cases like
this, with two imperative clauses in sequence, the hí clause expresses the action
necessary for the second one to take place. In other words, the usual causal value of
hí is found there as well, though the addition of imperative modality makes it
difficult to render in English.
I.15 Sequential deities (for the ṛtugrahas)
Although this hymn is in some ways a rote and formulaic listing of the
Ṛtugraha deities with invitations to drink of their respective cups, the poet does inject
some life in the hymn by varying the expected phraseology. After having established
the formula DRINK ṛtúnā in the first few verses, the poet introduces deviations from
that formula. In vs. 5 he urges Indra to drink not from the Brahman’s cup (as would
be standard: see II.36.5), but from the Brahman’s “largess” (rā́dhasaḥ) and ṛtū́m̐ ánu
substitutes for ṛtúnā (note the phonological crossing of únā: ánu). In vs. 6 there is no
invitation to drink, though ṛtúnā is found in another expression. In vs. 7 the expected
deity (“Wealth-giver”) appears in the nominative (draviṇodā́(ḥ)) as expected, but
there is no attached predicate: the verse goes off in a different direction. The Wealth-
15
giver is the subject of the next three verses (8-10) as well, but it is only in vs. 9 that
any drinking goes on. Here the imperative “drink!” is replaced by the desiderative
“desires to drink” (pipīṣati), and though there is an ablative of a priestly cup, it is one
belonging to a different priest and the verb used with it is not ‘drink’. Although this
is not high art, it does show that even the most cut-and-dried litany affords some
room to tinker with the verbal form.
I.15.1 The accent on píba is syntactically unnecessary and not well explained.
Oldenberg (ZDMG 60) suggests either that it's a not explicitly marked foundation
clause for the following clause, or that píba has a tendency to emphatic accentuation.
Although he does not want to get the accent from píbata ṛtúnā in the next verse,
where the accent is correct (following voc. márutaḥ; cf. also píba ṛtúnā in 4c, again
with correct accent), this seems a possibility. One might also note that in the second
Ṛtugraha hymn sequence (II.37.1-3), the verb is accented in the phrase píba ṛtúbhiḥ,
though again the accent is unnecessary. So perhaps píba ṛtúnā/ṛtúbhiḥ was a
separable refrain-phrase in the Ṛtugraha ritual, and therefore received accent even
when incorporated into a hymnal context. See now comm. ad III.32.1.
I.15.2: sudānavaḥ lacks accent and is therefore a vocative, not (as the tr. implies) a
predicative nominative. The predicated vocative has been much discussed in the lit.;
see Old, Noten ad loc. and Bloomfield, RR. On this repeated pāda see comm. ad
VIII.7.12.
I.15.3: Tvaṣṭar is called Neṣṭar (‘leader’) here because he regularly “leads” the wives
of the gods.
I.15.4: The three wombs are presumably the three ritual fires, so “at/by” would be a
more felicitous translation than the published “in.”
I.15.6: The voc. dhṛtavrata ‘of steadfast commandments’ is apparently a singular in
the Saṃhitā text, though the Pp. reads –ā. There is, however, no sandhi situation that
could have triggered a shortening of the final vowel. Although the epithet is several
times found in the du. modifying Mitra and Varuṇa (VIII.25.2, 8), it is more often in
the singular modifying only Varuṇa (e.g., I.44.14, 141.9), and this may have been the
intent here. The vrata is really Varuṇa’s province, not Mitra’s. However, the matter
is complicated by the fact that the clear dual dvandva voc. mítrāvaruṇa in b also has
a singular ending, rather than expected –ā.. I would attribute this shortening to an
attempt to match the short final of sg. dhṛtavrata, rather than assuming the
shortening affected both words for the same reason. Alternatively, as JL points out to
me, it would be possible to consider the final short vowel of dhṛtavrata an example
of Kuiper’s Law, with loss of final laryngeal in pausa, though this could not account
directly for the following mítrāvaruṇa, which is the real problem.
I.16 Indra
16
I.16.1: It’s not clear why Indra’s horses are called sū́racakṣas- ‘having the eye of the
sun’, a word otherwise (3x) used of gods. The awkward doubling of the enclitic tvā
(found in both a and c, as object of the same verb) might suggest that the c pāda was
borrowed from elsewhere. And indeed this hymn gives the impression of being
assembled from ready-made formulae; the proportion of repeated pādas is fairly high
(see Bloomfield, RR), not to mention sub-pāda repetitions. JL suggests, however, that
the repeated tvā might not be the result of careless doubling, but rather the stranded
object of a gapped repetition of vahantu in pāda a.
I.17 Indra and Varuṇa
I.17.3: The other standard translations take tarpayethām as a self-beneficial reflexive
“satisfy yourselves” with anukāmám referring to the gods’ desire (e.g., Ge “freuet
euch nach Herzenslust”; Re “Rassasiez-vous à votre gré”), but this doesn’t make
sense. The whole hymnic context depicts Indra and Varuṇa as givers, not takers, and
it’s not clear to me that gods ever desire wealth, per se. It is much more likely that
the poet is asking the gods to grant us wealth, and that the kā́ma is the mortals’, not
the gods’. For further discussion, with a strikingly parallel usage, see Jamison 1983:
140-41 and esp. n. 73, though I would not now emend the text to ánu kā́mam, as I
suggested there. Curiously, though Ge translates the verb as a reflexive, he goes to
elaborate pains to interpret the whole phrase as urging the gods to give us their
wealth, in other words much the same meaning as I favor.
In b rāyá ā́ is ambiguous as to case; it could be either dat. rā́ye or abl./gen.
rā́yas (so Pp.). Ge takes it as a gen., construed with anukāmám, but rā́ya ā́ is a fairly
well-established expression (e.g., I.81.7, III.56.6) and the rā́ya seems too distant from
anukāmám to be naturally construed with it. Most other tr. seems to favor the dat.
(see esp. disc. of Old ad loc, also Re.), but I weakly favor an abl. reading “from
(your) wealth.”
A further question is what noun to supply with nédiṣṭham ‘closest’ or what
else to do with it. Ge suggests ‘wealth’ in a note; Re seems to take the word as
adverbial (“de la manière la plus proche,” whatever that means). My supplied
“friendship” is based on two occurrences of ā́pyam (VII.15.1, VIII.73.6) and one of
āpím ‘friend’ with nédiṣṭham. Esp. telling is VIII.73.6 nédiṣṭhaṃ yāmy ā́pyam “I
beseech (you two) for closest friendship,” with the same verb as here.
I.17.4: A curious construction: bhū + GEN., which seems to express partitive value:
“become (part of) X” à “partake of,” though the path to partitive meaning is not
straightforward. (Other translators seem to feel less guilt about making this leap than
I do.) Closest to it formulaically is vidyā́ma + GEN, “might we know of X.” Cf.
vidyā́ma sumatīnā́m (I.4.3, X.89.17) “might we know (of) your favors.” The oddity
of the construction is exacerbated by the emphatic pāda-initial repetition (a, b) of
indeclinable yuvā́ku ‘of you two’.
17
I.17.5: As noted in the intro., this is the middle verse of the hymn and the only one in
which Indra and Varuṇa are separated, and in my view translations that don’t take
this into account are likely to be on the wrong track. (So, Ge’s “Indra, Varuṇa sind
der preisliche Rat der rühmlichen Tausendschenker,” which puts all nominatives and
all genitives together.) The rhetorical structure of ab, nom. sg. – gen. pl. / nom. sg. –
gen. pl., invites an association of each nom. sg. with its immediately following gen.
pl., producing a contrastive pair of Indra associated with thousand-giving and Varuṇa
with (something) to be proclaimed. I have followed this rhetoric clue and, further,
have tentatively supplied a noun (‘master’) to head the gen. phrase. Alternatively,
krátuḥ of c may be the head (“Indra becomes the krátu of …”), as WG take it (also
Ge).
There is the further question of what śaṃsyā̀nām refers to. I have somewhat
reluctantly supplied ‘riches’, since this is a theme of the hymn and śaṃsyà- several
times modifies rayí- and similar words (e.g., VIII.60.11, X.47.2). However, it also
regularly qualifies verbal products like ukthyà- or ukthá-, and given Varuṇa’s nature,
an association with “(solemn words) to be proclaimed” is probably more likely than
with “(riches) to be proclaimed.” I might therefore modify my published translation,
though the desire for wealth is quite strong in this hymn (see vss. 3-4, 6-7).
I.17.8: The doubled nū́ nú ‘now now, just now’ is found only here, though doubling
with an intercalated particle is found (nū́ cin nú I.120.2, VI.37.3, VII.22.8). It is
possible that the sequence nū́ nú vām is meant to evoke a form of √nu ‘bellow, shout’,
in this verse concerning the poet’s praise of the gods. Various forms of song and so
forth serve as subjects of √nu. Cf. nearby I.6.6, 7.1.
I.17.9: suṣṭutí- ‘lovely praise’ is not a particularly good obj. of huvé, which ordinarily
takes the addressee, not the content of the call. (See … vām … huvé in vs. 7). Here
the semantic disharmony may suggest that the lovely praise is personified and urged
to do her part to please Indra and Varuṇa.
I.18 Brahmaṇaspati and Sadaspati
I.18.1: Though somā́nam is assigned to a -man-stem by Debrunner (AiG II.2.760), it
seems preferable to analyze it, with Kuiper (IIJ 15 [1973]: 190–94 [my thanks to ET
for the ref.]), as having the so-called “Hoffmann suffix” (*-Hon-/ -Hn-) (Hoffmann
1955 = Aufs. II, 378-83) added to the thematic noun sóma-, of the same type as Aves.
mąϑrān- ‘possessing mąϑras’ to mąϑra-. Our somā́nam is a hapax, so there are no
diagnostic forms; a masc. agent noun in -mán would also have suffixal accent and a
long suffixal vowel (cf. dharmā́nam, -as ‘upholder(s)’). In favor of the Hoffmannsuffix interpretation is the quadrisyllabic scansion, inviting distraction of the ā-,
which should not occur in a man-stem. The accent might be a problem; Hoffmann is
somewhat cadgey about the accent of these forms (not difficult, since most of his
examples are Avestan), but he does suggest (p. 381) that the original accent of the
acc. sg. might fall on the suffix, as here. Kuiper makes no mention of the accent.
18
Most translators (Ge ‘Lautsingenden(?)’, Re, Schmidt B+I, WG) take
sváraṇam to the ‘sound’ root √svar, though Gr connects it with svàr- ‘sun’
(‘glanzreich’). The metrical distraction to suváraṇam might favor a connection with
the ‘sun’ word, since, as far as I know, the ‘sound’ root is never so distracted. This
hapax is also phonologically reminiscent of svàrṇara- ‘possessing solar glory’,
whose initial cluster is always distracted and which almost always occupies the end
of the pāda, as our word does. At the very least, it is likely that svàraṇa- is a pun.
It is surprising to find Kakṣīvant Auśija, the dazzling poet of I.116-26, in the
context of this rather simple and mundane hymn, for Medhātithi certainly lacks
Kakṣīvant’s skill. The request must then be seen as a species of wishful thinking. I
translate auśijá- literally, as ‘descendant of a fire-priest’, though it is also
Kakṣīvant’s patronymic, because I think the word previews the Agni theme of the
second half of this hymn. The publ. tr. follows Ge in assuming a desired
identification of Medhātithi with Kakṣīvant: Ge “Mach … zu einem (zweiten)
Kakṣīvat.” ET points out that there is actually no overt mention of Medhātithi and the
verse could simply mean “make Kakṣīvant a possessor of soma …” But then we still
must explain the presence of Kakṣīvant here.
I.18.3: The śáṃso áruruṣaḥ … mártyasya “(male)diction of the grudging mortal,”
which is dreaded by the poet, anticipates the desirable nárāśáṃsa- ‘praise of men’
that opens the last verse of the hymn (9).
I.18.6: ayāsiṣam: I take this form to the root √yā ‘beseech, implore’, not to √yā
‘drive, go’, which does, admittedly, have a well-attested –siṣ-aorist. So also WG,
though other translators (including Schmidt, B+I) render as a verb of motion (Ge
“habe ich … angegangen”). That interpretation isn’t impossible, but ‘beseech’ fits
the context better.
I.18.7: If the referent of yásmāt is Agni and the two halves of the hymn concern the
verbal and the physical parts of the sacrifice respectively, as I argue in the intro., this
verse makes particular sense: even a skilled poet has to get the oblations right.
I.18.8: All tr. take hótrā as nom. sg. and the subject of the clause. But since the point
of this verse and the last is the labor Agni expends in making the sacrifice succeed, it
seems better to keep him as subj. of gachati, as he was of ṛdhnoti (a) and kṛṇoti (b).
hótrā can then be an instr. sg. of the ā-stem, as commonly. The ritual model in which
Agni goes to the gods with the offering is of course quite widespread in the RV.
I.18.9: Note the high proportion of sibilants in this verse.
The meaning of makhá- and its relatives (here represented by the second
member –makhas-) is much discussed. I consider it to have both martial and
bountiful senses. I take the original sense to be martial and, despite some difficulties,
believe the often suggested connection with Grk. μάχομαι. The ‘bounteous’ sense
comes, in my opinion, from secondary association with maghá-, etc. In this
19
compound sádmamakhas- most tr. take the second member in a ‘give, be bounteous’
sense (e.g., Ge “der einen Wohnsitz beschert,” Re “qui confère-généreusement un
siège,” WG “den … mit einem Himmelssitz freigebigen”). However, the martial
sense fits the context well. The image is of the smoke of the ritual fire rising to, and
thus visually ‘besieging’, heaven itself (so Gr). See Old’s disc.; Scar (277)
questioningly suggests both.
I.19 Agni and the Maruts
I.19.4: Note the double etymological figures arkám ānr̥cúḥ ‘chanted the chant’ and
ugrā́ḥ ... ójasā ‘strong ones … through strength’, with the former nested inside the
latter. There is also phonological play between ...ám ānr̥- and ánādhr̥I.19.7–8: Ge (sim. Re) supplies a verb (“kommen”) with 7b. I am, as usual, reluctant
to do so, but as ET points out, swinging the mountains across the sea is a very
peculiar image. I would prefer to supply the ‘stretch’ (ā́ … tánvanti) of the next verse,
since the b pādas of 7–8 are structually identical, producing a tr. of 7ab “Those who
make the mountains swing, (as they stretch) across the undulating sea.” However,
Ge/Re don’t construe the ‘stretch’ verb with 8b either.
I.20 Ṛbhus
I.20.1 Calling the Ṛbhus “the godly breed” (devá- jánman-) is a bold way to begin
this hymn precisely because they did not begin as gods but obtained immortality by
their wondrous deeds – the deeds about to be recounted in the hymn.
I.20.2: Although vacoyújā is principally a dual acc., modifying the two fallow bays
(hárī) as usual, it could also secondarily be taken as an instr. sg., modifying mánasā
‘with mind’. In any case, ‘speech’ and ‘mind’ are implicitly contrasted in this verse,
and in addition mánasā contrasts with āsayā́ ‘by mouth’, referring to the means of
creating.
I.20.3: Though Ge (/WG) takes sabardúgha- as the cow’s name, the word has a
literal sense that works well in context.
I.20.5: A reference to the Third Pressing, with which the Ṛbhus are associated.
I.21 Indra and Agni
The verbal “hero” of this hymn is the nom.-acc. du. pronoun tā́ ‘these two’,
which represents the pairing of the otherwise very different gods Indra and Agni.
The form appears 6x in the hymn, with an additional gen.-loc. táyoḥ in 1b.
I.21.6: The opening of this vs., téna satyéna, is of course the standard signal of the
later truth formulation (satyakriyā). It is difficult to impose such a value here, but the
20
phrase can be seen as a summary of the praise-hymn that precedes this final vs. and
therefore as the grounds on which the poet asks for their vigilance and help.
Ge (followed by WG) considers the padá- that is to be watched over the track
or traces of something demonic or monstrous (“Die Spur der Unholde”), presumably
the rákṣas- of the previous vs. But Re (XIV.121) points out that the root √gṛ ‘be
watchful, wakeful’ always governs an object with favorable sense. Still, the exact
reference of padá- is unclear. As the word is often used of the ritual ground (e.g., in
the phrase iḷás padé “in the footprint of refreshment” [I.128.1, etc.]), I have
interpreted it in this way in the transl. But it’s also possible to invoke another
common use of the word, for the cosmogonic footsteps of Viṣṇu, an example of
which is in the last verse of the next hymn, also close to a form of √gṛ (I.22.21):
jāgr̥vā́ṃsaḥ ... víṣṇor yát paramám padám “watchful (over) what is the highest
footstep of Viṣṇu.” These two views can in fact be reconciled. As noted below, the
“highest footstep of Viṣṇu” in that verse is probably a reference to the sun, identified
with the ritual fire, and therefore here “the footprint of discernment” can be both the
ritual ground and Viṣṇu’s footprint in heaven.
I.22 Various divinities
I.22.4: The use of the diminutivizing -ka-suffix on dūraké ‘at a (little) distance’
reinforces the point of the verse, that it’s not much of a trip for the Aśvins to come
here.
I.22.6: It is not entirely clear why we “desire the commandments (vratā́ni)” of
Savitar, though I think it is likely that it is because his vrata keep the world
functioning, esp. the cyclical rising and setting of the sun.
I.22.10: Note the opening phonetic figure ā́ gnā́ agna.
The shadowy goddess figures mentioned here are difficult to sort out. I
assume that there are three goddesses here, with Hotrā Bhāratī the double name of
the deified libation belonging to the Bharatas. In this passage Ge takes them as two
different figures (also in II.1.11, III.62.3), but in I.142.9 as a doubly named single
figure. See Old (SBE, ad I.142.9): “Hotrâ Bhâratî, i.e., the personified Offering of
the Bharatas, seems to be one goddess, more usually called simply Bhâratî,” with reff.
I.22.11: The goddesses “with unclipped wings” (áchinnapatra-) are found only here
in the RV, but appear in a set of YV mantras used in the Agnicayana at the placing
and heating of the ukhā pot (for firing the bricks). See VS XI.61, MS II.76, KS
XVI.6, ŚB VI.5.4.8, etc.. A number of other female divinities are mentioned in the
same context. What áchinnapatra- actually refers to is unclear to me. Griffith (ad VS
XI.61) suggests “moving freely and without interruption.” I also do not see any sign
of the later Agnicayana employment here, simply an association with other female
divine figures.
21
I.22.13: mahī́ is grammatically ambiguous, as it can be either fem. singular or fem.
(or neut.) dual. Most tr. take it as the former, modifying only dyaúḥ, but all things
being equal, it is preferable to take dyauḥ as masculine (though fem. occasionally is
unavoidable), esp. as mahī́ by itself can refer to the earth without further
specification. I see no reason why mahī́ cannot be dual here, modifying the conjoined
NP dyaúḥ pṛthivī́ ca.
I.22.14: The meaning of this verse is entirely obscure to me, and it comes as a
surprise in hymn (or set of hymns) displaying no other verbal tricks at all. (It is true
that the Gandharva spreads obscurity almost every place he appears in the RV.) I
might speculate that what the verse is conveying is that by their inspired hymns the
poets have achieved a place in the highest firmament, where the Gandharva often
seems to be located, and where they consume the luscious food produced by Heaven
and Earth. If this speculation has any merit, the instr. dhītíbhiḥ should not be taken as
an instrument of licking (“lick … with their poetic insights”), but rather as an instr.
of cause (“by reason of their [prior and successful] poetic insights”), so the publ. tr.
could be amended accordingly.
I.22.15: On anṛkṣará- as ‘not sweeping men away’, rather than ‘thornless’, see
Jamison 1993 (Fs. Rix). The standard ‘thornless’ interpretation has no real support; I
derive the word instead from a-nṛ-kṣara-, to the root √kṣar ‘flow, stream’, cognate
with Greek φϑείρω ‘I destroy’. Formulaic and textual support for this analysis is
given in the art. cit.
“A place to settle down” (nivéśanī) might have been better tr. adjectivally
“bring to rest, causing to settle down,” but this causes awkwardness in the English.
I.22.16-18: This tṛca is marked by minor ring composition: the átaḥ [sandhi áto]
‘from there’ that begins 16a also begins the last pāda of 18, in the same sandhi form.
I.22.20: The image in pāda c needs a little unpacking. The “eye in heaven” is of
course the sun; the incongruity is that it is depicted as “stretched out,” which might
suggest an elongated ovoid sun, not a happy picture. The concentrated expression of
pāda c rests on the common formula of the sun stretching (through space) with its
rays (raśmí-), as in VII.47.4 yā́ḥ sū́ryo raśmíbhir ātatā́na “towards which [=waters]
the sun has stretched with its rays.” Note that sūráyaḥ ‘patrons’ ending b
phonologically evokes the absent sū́riya- in c.
I.22.21: This verse also contains some semantic incongruities: “kindling the footstep”
is of course a curious expression, and that poets, presumably human, perform this
action on the “highest footstep of Viṣṇu,” usually an expression for highest heaven,
makes it all the odder. Ge (also Old, WG) plausibly identify “the highest footstep of
Viṣṇu” here as the sun, and the verse would therefore express the common notion
that priests kindling the ritual fire at dawn cause the sun to shine.
22
I.23 Various gods (Praügaśastra)
I.23.4-6: This tṛca to Mitra and Varuṇa stations the two names in three different ways
in the three verses. In 4 mitrá- opens the first pāda and váruṇa- the second; in 5 they
are expressed in the dual dvandva mitrā́váruṇā in c; in 6 they again open the first two
pādas, but with váruṇa- first and mitrá- second.
I.23.8: It is striking that all of ab is made up of vocatives, with each one accented:
because of the place of the accent, three of the four words have to be vocatives; only
índrajyeṣṭhā(h) could be nom. pl., since índra- has inherent initial accent. The first
word of c, víśve, may also be and probably is a voc.
It would be satisfying to have three GOD-X bahuvrīhis parallel in semantic and
morphological structure, but although Ge (followed by WG) interprets the rāti- in
pū́ṣa-rātayaḥ as an agent noun (“mit Pūṣan als Gönner”), rātí- both as simplex and in
its fairly numerous compounds is always an abstract ‘giving’ or concretized version
thereof (‘gift’), as its morphology as a deverbal feminine abstract would require. I
therefore take the third cmpd as a tatpuruṣa pūṣa-rātí- the gift(s)/giving of Pūṣan’. It
owes its initial accent to the fact that it is a voc. (as does the bahuvrīhi márudgaṇāḥ
[expect *marúdgaṇa-].) The ‘gift’ or ‘giving’ of Pūṣan is described as auspicious
(bhadrā́) in VI.58.1.
I.23.9: Ge (WG), following Gr, tr. sáhasā as an adj. modifying Indra (“mit dem
starken Indra”), but this is of course impossible. With that accent it must be a neut.;
accent shift to *sahás- would turn it into an adjective, but this form is unattested (vs.,
e.g., yáśas- ‘glory’à yaśás- ‘glorious’).
On the analysis of fairly common 3rd sg. īśata (always with mā́(kiḥ)) as an
injunc. to a reduplicated aor., see Hoffmann 1967 (Injunk.): 64-66. Although
semantic and syntactic factors impose this interpretation, it nonetheless does not
seem entirely satisfying, because the thematic stem envisioned (īśa- [never
accented]) seems insufficiently distinguished from the presential perfect stem ī́ś-, and
Hoffmann does not suggest a mechanism for deriving such an aorist stem from either
the presential perfect stem or directly from the root.
I.23.13-15: As noted in the intro., I take the “king” that Pūṣan is returning as Agni,
not, with most tr., Soma. The theme of the finding and returning of the god-in-hiding
in 14 is much more appropriate for Agni than for Soma, and the “glitttering barhis”
is also more likely to be associated with Agni. Both Agni and Soma are called
“buttress, support” (13) (Agni in V.15.1–2) and both are called king (14). The drops
(índu-) in 15 might seem rather to indicate Soma, but in VI.16.16 Agni is
strengthened by drops (índubhiḥ), which must be drops of ghee.
I.23.15: The ritual referent of the “six yoked” entities (ṣáḍ yuktā́n) is unclear whether
Soma or Agni is taken as the main referent: those in the Soma camp consider the
yoked ones to be the dakṣiṇās due the poet; I consider them the flames of the
23
recovered Agni, roused by the drops of ghee. In terms of the simile in c, the six
yoked ones are clearly the teams for plowing. See esp. AV VI.91.1, cited by Schaefer
(1994: 197). On anu √sidh as ‘entlangtreiben’, see Schaefer 1994: 196–97.
In my translation I have reversed the renderings of the finite intensive
(unaccented carkṛśat in c) and the intensive participle (accented anuséṣidhat) to
make the relationship between simile and frame clearer.
I.23.16: Contra Ge and others but with Re, I construe the gen. pl. adhvarīyatā́m ‘of
those performing the rites’ with ádhvabhiḥ ‘along the roads’, as I do not think the
waters are the mothers of the priests (Ge “Die Mütter der diensttuenden (Priester)”).
I.23.18: Ge and others attach b to a, with c a separate sentence, which is also possible.
I.23.19: Again I disagree with most tr. in the disposition of the b pāda. I take it with a,
because I think we mortals are to glorify the waters. Others think the gods should
become strong or victorious in order to praise the waters – but gods are not usually
the agents of such praising.
I.23.20-23 = X.9.6-9, save that the fourth pāda of 20 is missing in X.9.6.
I.23.20: The semi-direct speech and esp. the casual conversational tone of this verse
are quite unusual for the RV. Although most tr. take the whole as indirect speech
(“Soma said to me that…”), the construction of the verse seems to me to mix direct
and indirect speech, with c, couched in the acc., shading into indirect speech. I take
the final pāda separately, as it is absent from the parallel in X.9.6.
Note the distraction of the apsú antár phrase that opens 19a into two pāda
opening words: 20a #apsú … b #antár
I.23.23: Most tr. take ā́paḥ as acc. here (“I have followed the waters.”). This is of
course possible: the nom. of this stem does get used for acc. occasionally in the RV.
However, such an interpr. is not nec. in this case, since it is easy to supply an acc.
“you.” And the fact that the two previous vss. (21a and 22a) contain vocatives ā́paḥ
and āpaḥ respectively supports a vocative interpretation here.
I.23.24: It is difficult to know what, if anything, is the referent of asya. I have
tentatively supplied ‘hymn’, but the poet may simply be calling upon the gods to
witness the general situation (so Ge). The same expression √VID me asya is found in
the refrain of I.105, except with accented asyá, for which reason I more confidently
supply ‘speech’ there.
[I.24–25 JPB]
I.26 Agni
24
I.26.1: As often, the hí in the first of two imperative clauses marks the action urged
in ab as subordinate to and the basis for the consequence expressed in c. See
Brereton 2012.
I.26.2: As on some other occasions the pāda-final vácaḥ, superficially a nom.-acc. (or
a stem form), is to be construed as an instr. with the instr. adj. (divítmatā in this case),
whatever the source of this truncated form. (See Hale [Fs. Melchert] for an attempt,
not entirely successful in my view, to see it as an archaic zero-grade s-stem instr. [*es-H, not the renewed and standard *-es-eH].)
I.26.3: A paradigm ex. of the use of the act. of √yaj to express sacrificing on behalf
of someone other than the grammatical subj.
I.26.5: In the phrase … asyá naḥ, mándasva sakhyásya ca most tr. render the ca as
‘also’ and construe the demonstrative with the noun (Ge “… freue dich auch dieser
Freundschaft mit uns”). However, the most natural way to take the syntax is as a
conjoined NP: “of this (x) and the comradeship of ours.” The question is what asyá is
modifying. Following Gr I supply ‘ritual grass’, because barhíḥ is the most recent
ritual referent (4a). Although “rejoicing in ritual grass” may seem an odd activity, cf.
VIII.13.4=VIII.15.5 mandānó asyá barhíṣo ví rājasi, with the same verb √mand
‘rejoice’ and the fuller version of the nominal phrase. (As it happens, in that passage
I construe the barhis phrase with the main verb ví rājasi, but it is certainly
ambiguous.)
I.26.8: Although most tr. implicitly take pāda a as subordinated to b (e.g., Ge. “Denn
wenn …, so…”), again the ca suggests rather that they are coordinate structures, with
pāda a a nominal sentence (svagnáyaḥ predicated of devā́saḥ) coordinated with the
finite verb dadhiré in b. This verb is accented because it is under the domain of hí in
a. In this reading, pāda c functions as the main cl. of ab.
I also supply ‘him’ (=Agni) as first object of dadhiré, again unlike most tr.
(eg., Ge “so bringen sie auch uns Erwünschtes”). For the reasons for this in the
structure of the hymn, see disc. in intro.
I.27 Agni
I.27.1: vandádhyai is a predicated infinitive, serving as the main verb.
Pāda a contains a likely pun, the possessive –vant-stem adj. vā́ravantam
modifying Agni. In the first instance it means ‘having (that is, providing) choice
things’ (vā́ra-, to √vṛ ‘choose’), but the homonym vā́ra- means ‘hair, tail-hair’ (cf.,
e.g., nearby I.32.12) and so the -vant-stem can mean ‘possessing (long) tail-hair’.
Either of these meanings is applicable to the horse of the simile, which has tail-hair
by nature and brings choice things by winning races and contests. ‘Bringing choice
things’ is more applicable to Agni than ‘having hair’, but his flames could be so
25
conceived. Note that Agni is characterized as “ghee-haired” (ghṛtá-keśa-) in
VIII.60.2.
I.27.2: With Old I supply śávasaḥ with sūnúḥ, to complete the common epithet of
Agni “son of strength,” which is suggested by the associated instr. śávasā. Indirect
support might come from the parallel voc. sahaso yaho ‘o young (son) of strength’
(with a different word for strength, also regularly appearing as an epithet of Agni)
that ends the preceding hymn (I.26.10c). However, this supplement is not strictly
necessary, and most tr. do not supply it (e.g., Ge simply “unser Sohn”). In favor of a
translation like Ge’s is the presence of naḥ in pāda a. However, this could simply be
a Wackernagel place-holder for asmā́kam in c. Although the difference might seem
slight, in fact the two interpretations are quite different. Ge’s emphasizes that we
have created Agni, supplying ‘our’ with śávasā (“unser Sohn durch (unsere) Kraft”);
this might suggest that Agni is in our debt because we begot him and he ought
therefore to become our benefactor. The other tr. invest Agni with his own strength
and hopes that he will graciously use some of it to benefit us.
I.27.6: With Ge (/WG) and Re I take the expression in pāda b as a simile or
comparison that provides the basis for the verb ‘stream’ (kṣarasi) in c. Agni/the fire
cannot literally be on or in a river’s wave.
I.27.7: It is difficult not to interpret the agent noun yantā́ in c as a periphrastic future,
given the subjunctives in ab, pace Tichy 1995: 226. Although it is sometimes
claimed that no examples (or “no certain examples”) of this usage are found in the
Saṃhitās (see esp. Macdonell VGS §152), there are too many passages in the RV
where a future interpretation of the –tar-stem is more natural and fits the context
better than a purely agentive one.
I.27.8: Since this verse continues the thought of 7, the agent noun paryetā́ in b should
likewise be future in value, despite Ge’s “Keiner überholt ihn.”
asya … káyasya cit: most tr. take these two genitives as coreferential, with the
indefinite referring to a person – so Ge’s “Keiner überholt ihn, wer er auch sei.” But
the person in question has already been defined as a client of Agni’s, and so an
indefinite seems odd in context. Moreover, the other two instances of paryetár- both
take inanimate complements; cf. VII.40.3 ná tásya rāyáḥ paryetā́sti "There exists no
one who can encompass his wealth." Therefore I take káyasya cit as referring to
anything belonging to the favored man, expressed by asya. Ge’s note seems to lean
in this direction, but not his tr.
On śravā́yya- see I.31.5 below.
I.27.10: The supposed voc. járābodha in a is problematic on several grounds. It is
generally taken to mean something like “attentive to the early call,” but 1) the first
member, jarā́-, is only attested in the meaning ‘old age’ (hence Old’s suggested “im
(bis zum) Greisenalter wachend”), and 2) the second member, the thematic nominal
26
bodhá-, is not otherwise found in the RV (though it does occur in the AV). I
therefore prefer to interpret it as a pair of linked imperatives, járā bodha. The latter
is found 5x in the RV, as impv. to the thematized aor. to √budh ‘awake’; the former
would belong to the thematic pres. jára- of √gṛ/jṛ ‘sing’. Although this present is
ordinarily only middle (vs. járati ‘make old’), attraction in voice would not be
surprising in a construction like this. The long -ā of járā simply reflects the common
lengthening of the 2nd sg. impv. Although we might expect bodha to be accented, it
may have lost the accent when the construction ceased to be understood, or it may
never have received it in this close semantic nexus. Under this analysis the order of
imperatives is actually “sing (and) awake!” which I have reordered for clarity.
Alternatively, we might take járā to √gṛ/jṛ ‘awake’, which likewise builds a Class I
middle pres., and translate “awake (and) take heed.” See Gotō 1987: 150-56 for
discrimination of the various Class I jára- presents.
Since dṛ́śīka- is otherwise a neut. substantive, I have rendered it as
appositional to stómam rather than as an adjective (e.g., Ge “ein schönes Lied”). So
also Re (“un corps-de-louange, chose belle à voir”). Note also the synesthesia, with
the praise-song something to be seen, not heard.
I.28 Indra (Abbreviated Soma Rite)
See intro. for detailed discussion of my interpretation of the context of the
hymn, which differs significantly from the standard view.
I.28.1: Both pṛthúbudhna- ‘having a broad bottom’ and ūrdhvá- appear to be doubleentendres. The salacious references continue, more clearly, in vss. 3-4.
I.28.6: The reference in this verse is not entirely clear, but “the lord of the wood”
(vánaspáti-, ordinarily a word for tree, or an esp. prominent tree, then applied the
sacrificial post) is probably here the pestle and metaphorically the erect penis. If so,
ágram might be better tr. ‘tip’ than ‘top’ and the whole clause as “the wind blows
across [rather than ‘through’] your tip.” In an unpublished paper on this hymn,
"Rgveda I.28 and the Two Forms of Pressing Soma," Hanns-Peter Schmidt cites a
Kuntāpa verse AV 20.136.6 = ŚŚS 12.24.2.7, which he tr. "The harlot, stepping over
the mortar, said, 'Just as on thee, o tree (o wooden mortar), they strike (with the
pestle) so may they strike on me,'" which also shows the connection of mortar and
pestle with unlicensed sex. If the tr. is correct, the “lord of the wood” there is the
mortar, not the pestle, as it seems to be here. In vs. 8 below, both the mortar and the
pestle seem to be so called.
I.28.7: On the motions involved, see Schaefer 1994: 163-64.
I.29 Indra
I.29.2: Ge takes … táva daṃsánā as a nominal sentence “du hast ja die
Machtvollkommenheit.” This is possible. But with Gr and Re, I prefer to take
27
daṃsánā as an instr. (as it often is), in order to allow the whole verse to be a single
sentence.
I.30 Indra, Aśvins, Dawn
I.30.1: The word krívi- makes trouble wherever it shows up. In some of its
occurrences it appears to be a personal name, but that is unlikely here. It is not even
clear if all the occurrences of krívi-, even in non-personal uses, belong to a single
stem. Here the context favors an equine reference, and I have taken it as designating
a particular color of horse, viz., ‘blood-red’. Assuming, with most scholars, that
krívi- in at least some of its usages is connected with krívis- in the hapax adj. krívirdant- (I.166.6), which possibly means ‘having gory teeth’, and that both are
etymologically connected with kravís- ‘raw, bloody flesh’, as a color term ‘blood-red’
would work well. The details of the derivational relation I leave to others, but a
putative -i-stem to the underlying root krū < *kruh2 (in krūrá- [AV+]) would have
the shape *kruh2-i- > *kruv-i-, and assimilation to krívi- would not be difficult.
The verse is structured by number disharmony — the pres. part. vājayántaḥ is
nom. plural, but the finite verb siñce is singular — reflecting the common conceptual
fluctuation between the collectivity of priests and singers and the priest-poet
speaking in his individual voice.
I.30.2: The number disharmony continues here, at least in my view. Ge. takes the sg.
rel. pron. yáḥ as referring to Indra and then supplies the verb ‘drinks’ to governing
the ‘hundred’ and ‘thousand’ phrases referring to soma. But if yáḥ is instead taken as
referring to soma and coreferential with the (unexpressed) subj. of the verb in the
main clause rīyate, no material needs to be supplied. Instead the singular ‘which
(soma)’ is defined as consisting of a hundred or a thousand separate drinks. So also
Re.
I.30.3: The form śuṣmíṇa (in sandhi) can represent either śuṣmíne (dat. sg.) or
śuṣmíṇaḥ (in turn, either gen.-abl. sg. or nom.-acc. pl.). (The Pp. reads –e.) Any of
these possibilities is possible in context, and so it may well be that the ambiguity is
meant. As a nom. pl. it could characterize the subj. (‘they’ = soma drinks), as Re.
takes it. As a gen. sg. it could refer to Indra, who is indeed regularly modified by this
adj. As a dat. sg. it could modify mádāya (so Ge), or it might still refer to Indra, in
well-known double dative construction of the type “for the tempestuous one for his
exhilaration” à “to exhilarate the tempestuous one.” I favor an interpretation that
attributes the word to Indra; among other things, this makes the unaccented asya in b
easier to account for. Preferably genitive śuṣmíṇaḥ, though I have not been able to
find a parallel expression.
On the surface, pāda a lacks a main verb, but it is possible that it is lurking
there. The subordinator yád appears as yán in sandhi before a nasal. The otherwise
unattested 3rd pl. injunctive of √i ‘go’ would have the same shape and is the expected
28
verb in this idiom (cf., e.g., 3rd pl. impf. sám āyan X.27.8). I therefore suggest we
have a notional haplology sáṃ *yád yán.
I do not understand the function of hí in b, which triggers accentuation of the
main verb dadhé in c. I would at all costs prefer to avoid attributing pure emphatic
function to hí, but this passage comes perilously close to that.
I.30.4: ohase is generally taken as a 2nd sg. mid. to the root √ūh / oh, which has a
messy set of forms. Cf., e.g., Ge. “Diese Rede von uns weisst du gewiss zu
würdigen.” However, I interpret it as a 1st ps. –se form of the type stuṣé ‘I shall
praise’, gṛṇīṣé ‘I shall sing’, all of which fall into this semantic sphere. Indeed the
root has a –se formation of different shape in VIII.5.3 vā́cam … ohiṣe, with almost
identical object (our passage: vácaḥ). For the thematic/(pseudo-)subjunctive form
here, cf. arcase ‘I shall address” (X.64.3). Despite the complications involved in
positing a second 1st ps. -se form to this root and separating ohase from the identical
form in VIII.80.9, which I take as a 2nd sg. in passive usage, I prefer my
interpretation to that of Ge (/WG) ‘value’, Re ‘take into consideration, take note of’,
because these latter seriously attenuate or alter the meaning of the root, which
otherwise means ‘solemnly proclaim, praise, vaunt oneself’.
I.30.5: The word order of ab is rather tortured. The clause is simply a nominal
sentence forming a relative clause, but the relative pronoun phrase yásya te, instead
of appearing in 2nd or modified position after stotrám as is overwhelmingly common
(see, e.g., modified 2nd in 2a śatáṃ vā yáḥ … above), comes at the very end of the
hemistich, separated from its noun by three vocative phrases. This still counts as 2nd
position in some sense, but the poet is pushing the limits. The te is simply there to
indicate that the rel. pron. has 2nd ps. reference, which is of course not syntactically
coded onto the relative. Cf. the common phrase táṃ tvā (e.g., 10a below), where the
2nd ps. pronoun has the same function: to give 2nd ps. ref. to the demonstrative.
I.30.8: yádi here stands for yád *ī, ‘when … it’ rather than ‘if’. The pronominal
enclitic ī has been shortened before the initial cluster of śrávat. See Jamison 2002.
I.30.9: Although huvé in c is morphologically problematic, its interpretation is
imposed by context: a past-tense 3rd ps. ‘he called’. But this is the only 3rd ps. huvé
(in contast to over 70 exx. of 1st ps. huvé, -e), and it must be preterital not present, as
huvé otherwise is. I have no explanation.
The referent of the te ‘your’ in the same pāda is unclear. Assuming the
relevant constituent is “your father’ (te … pitā́), te should refer to the poet, or some
poet, in whose lineage “you” are, but I would expect the poet to be the 1st ps. speaker
of the first huvé (pāda b). Perhaps the reference to the “ancient house” in pāda a
indicates that another, more venerable poet is on the scene, whose model (and that of
his forebears) the current poet is following. The tr. of pāda a supplies “your” with
“ancient house” – implying that the current poet is modeling himself on “you” and
“your father,” but it should be remembered that the “your” of a is not explicit in the
29
text. It could well be “my ancient house” (or indeed someone else’s). Though the
pāda is repeated in VIII.69.18, that passage is not helpful in interpreting this one.
However, I.87.5 contains an occurrence of pratná- modifying father that seems to
assert a similar entitlement to poetic speech by lineage: I.87.5 pitúḥ pratnásya
jánmanā vadāmasi “We speak by virtue of our kinship with our primordial father.”
I.30.11: Though it modifies asmā́kam ‘of us’ and “we” are definitely male,
śipriṇīnām ‘of the (well-)lipped ones’ is a feminine gen. pl. The form must be
contextually motivated: the other two pādas end with (masc.) gen. pls. in -Vnām / Vnaam (somapā́vanām [my preference, for HvN somapā́vnãm], sákhīnãm). Esp. the
latter, with -īnãm, invites a type of dittography: *śipriṇãm à śipriṇīnãm. But the
problem probably can’t be separated from that of equally unexpected fem. for masc.
návyasīnãm V.53.10, 58.1.
I.30.12: Ge (WG) take iṣṭáye as “dass du rasch kommst,” without comment, but it is
not clear to me what root they are using to produce ‘come quick’. There are several
roots √iṣ to which iṣṭáye could belong: ‘seek, desire’ and ‘send’ are the most likely,
along with the marginal root ‘prosper’ found in iṣáyati. A zero-grade of √yaj
‘sacrifice’ could (and does) also produce iṣṭí-. None of these comes close to ‘come
quick’; my conjecture is that they are connecting it with ‘send’, but forms of this root
are always transitive. Re more reasonably assigns it to the ‘seek’ root: “… que (tu)
cherches (des biens pour nous),” but must supply much material for it to work. I
suggest that it belongs to this same root, but in the sense ‘desire, want’, and that the
message here is the mutually reinforcing “we want you to want what we want.” This
expression is somewhat reminiscent of sá naḥ sanitā́ sanáye in vs. 16: “he the winner,
for us to win it,” though using two different but synonymous roots, rather than the
etymological figure of 16.
I.30.13: A noun needs to be supplied with revátīḥ ‘rich (fem.)’. Old suggests íṣaḥ
‘refreshments’, which works formulaically with the rest of the lexicon in the passage
(including the verb in c; cf., e.g., VII.64.3 iṣā́ madema, with an instr. as in our yā́bhir
madema). Ge’s “Geschenke” (gifts) is not so happy.
sadhamā́da out of sandhi could end in either -e (so Pp., followed by most) or aḥ. Complicating the decision is the fact that both a thematic stem sadhamā́da- and a
root noun sadhamā́d- are well attested. Though most tr. take presumed underlying
sadhamā́de as loc. to the thematic stem (Ge “bei dem Mahlgenossen”), I think it
possible that it is the dat. of purpose to the root noun. It is possible that Re’s “pour le
symposion” also assumes such a dative.
I.30.14: Vss. 14–15 are parallel in structure and phraseology, but this does not help
as much as we might like. To begin at the beginning, it seems odd to refer to Indra as
“one like you” (tvā́vān), and grammatically it is also problematic, because tvā́vantordinarily takes a 3rd ps. verb, not 2nd ps. as here. It is tempting to follow Ge’s path
and make ab into a subordinate clause (“Wenn einer wie du…”), but this doesn’t
30
work because the initial ā́ of pāda a must be construed with the ṛṇóḥ of c. The rest of
the first hemistich, two participial phrases, is somewhat awkwardly phrased (at least
in translation) but comprehensible: the praisers obtain Indra (or his like) in his own
person and implore him for aid, which he provides in c. That pāda uses an idiomatic
expression for fitting an axle between two wheels. As the presence of ná in 14c and
the structure of vs. 15 show, this expression is a simile, to which some other action of
Indra is being compared, even though in 14 there is no obvious frame. The meaning
of ā́ ṛṇóḥ in the frame is somewhat different from the idiomatic usage, but not, to my
mind, the “untranslatable wordplay” that Ge sees. I take it mean ‘fit out’, namely
‘provide’. In 14c I supply ‘refreshments’ from the preceding vs. (13), the first vs. of
the tṛca. Nonetheless, this translation of the tṛca, esp. 14–15, strikes me as less than
satisfactory.
I.30.17: śávīra-. Although this may just have a suffix –īra-, it is tempting to see it as
a bahuvrīhi containing a Caland form śavi- (root śū; cf. śū́ra-) plus írā- ‘refreshment’,
so ‘possessing powerful refreshment’. For the accent, see the many compounds with
śuci-, e.g., śúci-jihva- ‘having a blazing tongue’. If this is correct, the translation of
the same word in I.3.2 should be modified. There it modifies dhī́- ‘insight’, which
might not be as apt, but thoughts and hymns that provide refreshment are not foreign
to the RVic conceptual universe. Although we might expect *śuvi-, ET suggests
comparing presumed Caland forms RV ákravihasta- and PN dabhī́ ti-, which
apparently show full grade of the root.
The neut. adjectives gómat and híraṇyavat are generally taken to refer directly
to the gifts that the Aśvins will provide us (e.g., Ge “Rinderbesitz … Goldbesitz (sie
uns)!” But (as Ge suggests in his note) in I.92.16 and VIII.22.17, where the pāda is
repeated, the adjectives modify vártiḥ ‘circuit, course’, in turn the object of forms of
√yā ‘drive’ in VIII.22.17 (and often elsewhere). Since 17b contains yātam, I supply
vártiḥ here as well.
I.30.18: Because yójana- has a number of different meanings (‘yoking, stage of
journey, distance, route’), the compound samānáyojana- has received a number of
different translations. The analytic phrase samānéna yójanena occurs in I.92.3 of
Dawn’s journey, where again ‘route’ best fits the context.
I.30.19: Other passages also depict the Aśvins’ chariot with one wheel at rest
(presumably on earth, on the ritual ground) and the other in motion (V.73.3,
VIII.22.4), as Ge explains in his note.
The “head of the inviolable (bull)” is somewhat opaque, but is probably a
mystical expression for the ritual ground – more clearly expressed in mūrdhán
yajñásya “on the head of the sacrifice” (II.3.2, IX.17.6). Others have speculated that
the whole complex (the two wheels and the bull’s head) refer to a constellation. See
WG n.
I.30.20: “Fair-weather friend” translates the voc. kadhapriye and is an attempt at an
idiomatic English rendering of ‘when-friend’, i.e., ‘uncertain, fickle, or capricious
31
friend’ – following Bloomfield’s (RR, ad I.38.1) attractive explanation of closely
related kadhaprī́- as built to the Aśvins’ epithet adhapriyā ‘then-friends’. The epithet
is appropriate to the usually reliable Dawn in this verse because the poet questions
her as to where she will bestow her presence and favors.
I.30.21: All of pāda c is a vocative, though it is syntactically peculiar for the noun in
the simile to be in the vocative: áśve ná, literally “like o mare.” It must have been
attracted into the voc. by the voc. adjectives that are the points of comparison: citre
aruṣi “bring and ruddy.” See also I.57.3 below.
I.31 Agni
I.31.1-2: Some reciprocal vocabulary: in 1c the kavis (the Maruts) are born under the
vrata of Agni, while in 2b Agni as kavi tends to the vrata of the gods. In 2c note the
phonetic figure vibhúr víśvasmai bhúvanāya.
I.31.3: The mythology behind ab is related in I.143.2, as Ge points out; see also
VI.8.4. It is not clear here why a mythological situation should be couched in the
imperative (“become manifest!”), esp. as the second hemistich contains two
augmented imperfects.
I.31.4: Purūravas is found only here in the RV outside of his dialogue with the
Apsaras Urvaśī in X.95. It’s not entirely clear why he is here, but his is a speaking
name (“having much roaring”), which fits with the bellowing in pāda a. Moreover his
son Āyu figures several times in this hymn (vss. 2, 11), and Iḍā, said to be
Purūravas’s mother in X.95.18, also appears in vs. 11.
Contrary to most tr., I have segregated sukṛ́te sukṛ́ttaraḥ as a separate clause,
since this expression is found in different context in I.156.5. sukṛ́t- is usually used of
the priest or ritual performer, and this interpretation fits with the 2nd hemistich, which
refers to the ritual here and now. Starting a new sentence in pāda b also helps
mitigate the tense disharmony in this verse, where the present in the subordinate
clause (múcyase) is by most accounts connected to an imperfect in the main clause of
d (anayan), the action of which should temporally follow the verb in c (or to the
imperfect avāśayaḥ in a). By connection c instead with the last part of b, this
temporal reversal is avoided
Ge (following Gr) tr. śvātra- as simply “Kraft,” but the noun belongs with the
root √śū ‘swell’, and the image in this pāda is a vivid one: Agni’s “swelling” in his
two parents refers to the flaming up of the fire created by the kindling sticks (already
referred to in 2d). Both Re and WG render the ‘swell’ sense.
I.31.5: údyataśruce … śravā́yyaḥ: Ge’s (/WG’s) “ruhmbringend” is not possible for
śravā́yya-, because this is a (pseudo-)gerundive and hence passive. The dat.
údyataśruce can serve as agent, because RVic gerundives can take dat. agents (see
Jamison 1979 [Die Sprache 25] 137–38 n. 14), as also in other IE languages.
32
The rel. pronoun yáḥ in c has double reference: in c it refers to the priest in
the 3 ps. and is the subject of accented védā; in d it refers to Agni in the 2nd ps. and
is the subject of the accented āvívāsasi. This grammatical play cannot be easily
rendered in English.
ékāyu- is a hapax and also probably a pun. On the one hand it is formed like
viśvā́yu- ‘having a complete lifetime’, dīrghā́yu- ‘having a long lifetime’, hence
‘having a single lifetime’; on the other, it can contain the proper name of Āyu, who is
found in vss. 2 and 11 of the hymn, hence Ge’s tr. “im Alleinbesitz des Āyu,” WG
“den Āyu (als) einzigen (Opferer) hatte.” Although Ge (/WG) do not recognize the
“lifetime” possibility, the parallel formations make such a reading hard to avoid. In
the lifetime meaning, the compound presumably refers to the ritual fire’s duration
through a single sacrifice or, more likely, through the lifetime of the sacrificer. In the
personal-name reading, it would mean that Agni and our ritual fire have the ur-Aryan
sacrificer and clanlord (see 11b) Āyu on their side. In vs. 11 Agni is identified with
Āyu the clanlord, and so here, in that identify, he brings together the víśaḥ ‘clans’.
The publ. tr. does not register the personal-name reading, which should be remedied.
ágre: Although in absolute usage this word can refer to the beginning of time,
as it were, that sense seems unlikely here because of the present tense verb. I
therefore consider it to reflect a phrase like ágre yajñásya (VII.15.5), at least in the
primary reading. Secondarily, with the second meaning of ékāyu- (‘having Āyu as
yours alone’), it may allude to the primal sacrifice.
I added the parenthetical “divine” qualifying “clans” because gods (or their
qualities) are the usual object of āvívāsa-, and I still think that is the more likely
interpretation. However, ET points out to me that, in light of vs. 11b tvā́m … devā́
akēṇvan náhuṣasya viśpátim “You … did the gods create as the clan-lord of Nahuṣa,”
it might instead refer to human clans. Proferes (2007) argues at length, though not
mentioning this hymn, that a royal fire can unite several clans.
rd
I.31.6: sákman is a hapax, rendered by Ge as “in der Freundschaft” but by Re as “en
détresse” (followed explicitly by WG “aus der Not”). The former interpretation is
surely the correct one, taking it as a derivative of the very common root √sac
‘accompany’ and directly cognate with Aves (O+Y) haxman- ‘association,
community’. (So EWA s.v.) Re suggests a derivation rather from √sac2 ‘dry up’, but
the semantics of the derivation are not straightforward, and such a connection
requires both the apparent Avestan cognate and RVic hapax sákmya- ‘comradeship’
to be decoupled from sákman-. It is hard to see the motivation for this proposal. It
must have been that the context seemed not to be one of comradeship, but in fact
there is no problem with that meaning in context: Agni demonstrates his
comradeship with the man at issue by giving him aid.
I interpret the first hemistich to be entirely concerned with the sacrifice, as the
loc. vidáthe ‘at the ceremony’ suggests. In my view vṛjiná-vartani- ‘having twisting
turns’ envisions the sacrifice as a race-course, which the sacrificer must navigate, esp.
the tricky turn at the further end. That the sacrifice was commonly envisioned as a
course to traverse is clear from the word adhvara- ‘ceremony, rite’, related to
33
ádhvan- ‘road, way’ and found in expressions like rathī́r adhvarā́ṇām “charioteer of
the rites” (I.44.9, cf. VI.7.2). Although vṛjiná- ‘twisted, crooked’ often has a morally
negative sense (like its English glosses), in this case the twist(s) would simply be
tricky (that is, challenging) to steer. It is possible, with some other tr., to assume that
the twisting turns are not ritual but refer to unfortunate life circumstances, but then
vidáthe is hard to account for.
The image of the sacrifice as a race-course in ab then transitions to the image
of a race or contest in general in pāda c. The páritakmya- ‘crucial turn’ picks up the
vṛjinávartani- of pāda a. As so often, dháne is a truncated loc. absol. for the common
expression dháne hité (I.40.2, etc.) “when the stake is set.” The last pāda then turns
the contest image into an actual battle; the line between contest and battle is a thin
and permeable one in the RV.
I.31.7: Though both Ge and Re make amṛtatvé ‘immortality’ and śrávase
syntactically parallel, despite superficial appearances they are in different cases and
should be so rendered.
I.31.8: The question in ab is who is winning the stakes. I take it to be “us,” with our
winning enabled by Agni’s giving glory to our poet. Most take it to be the poet
himself, a poet identified as ours (“for our bard to win the stakes”). This is certainly
possible. Indeed Old (SBE) actually interprets sanáye … kṛṇuhi as a periphrastic
causative “make/cause to win,” with the poet the first object. Although Zehnder
(Periphrastic Kaus., 2011) does not discuss this passage, he does recognize (p. 61)
other examples of periphrastic causatives to √san ‘win’, which does not build a
morphological causative (expect *sānáyati, or possibly *sanáyati).
apásā návena: the suffixal accent of apásā should rule out a tr. ‘work’, but
most tr. ignore (Re actually cites it in his notes as ápas-) or overrule the accent. So
Ge “durch das neue Machwerk” (i.e., the hymn), sim. Re, WG. Although Gr cites
other forms of apás- in the meaning ‘work’, none of these is convincing. Therefore,
although it would simplify the tr. to take it as neut. ‘work’, the transmitted text can
only mean ‘worker, laborer’. My tr. implies that we have a new poet, or perhaps the
bard, made glorious by Agni in pāda b, takes a new lease on his poetic life.
Alternatively, we might follow Old (SBE), who alone paid attention to the accent and
tr. “with the help of the young active (Agni).”
I.31.9-10: Some patterned responsion in these two verses: 9 tanū-kṛ́t … prámatiḥ / 10
prámatiḥ … vayas-kṛ́t. However, although these two verses are roughly in the middle
of the hymn and patterned responsion is characteristic of omphalos verses, the
repetitions do not seem sufficiently important to constitute an omphalos. On the
relationship between prámati- and váyas- and between tanū- and váyas- see disc. ad
I.71.7, where it is suggested that tanū́- and váyas- are the tangible and intangible
elements that together make up a living being.
34
I.31.11: On Āyu and Iḍā see disc. ad vs. 4 above. This verse clearly refers to the
primal institution of the ritual fire at the Ārya sacrifice, but the details are somewhat
obscure.
Because of the tense differences between c and d (augmented impf. akṛṇvan
vs. pres. jā́yate), I follow Re in taking d as the content of Iḍā’s instruction, namely
that in mundane current-day terms Agni/ritual fire has a mere human father, the
sacrificer, though it was the gods who originally created him/it.
I.31.12: The theme of protection gets hammered home by the use of three different
roots in this semantic sphere: √pā (pāyúbhiḥ a), √rakṣ (rakṣa b, rakṣamāṇaḥ d), and
√trā (trātā́ c ).
I.31.13: The man who lacks even a quiver lacks weapons and is therefore defenseless.
My interpr. of cd differs significantly from the standard (though it is close to
Ge’s alternate, given at the end of his n. on 13cd). Most take yó rātáhavyaḥ as a
nominal rel. cl. referring to the worshipper, with coreferential resumptive pronoun
tám in the main clause vanóṣi tám, whose subject is Agni (e.g., Ge “Wer Opfer
spendet … den begehrst du.”). (Its initial position in its clause would of course
account for the accent on the main verb.) This interpretation has several merits: the
skeletal syntactic structure is clear; it explains the unusual position of the tám; it
parallels the structure of 14b; and rātáhavya- is more often used of mortals than of
gods (a usage that might be supported here by the parallel compound
práyatadakṣiṇa- in 15a). However, it makes the phrase kīréś cin mántram mánasā
impossible to construe: the acc. sg. mántram has nothing governing it, and the
“solution” in such tr. is simply to set it off by dashes and hope for the best. I
therefore prefer to take cd as a single relative clause with Agni as subject. The adj.
rātáhavya- can modify gods, including Agni in IV.7.7. This reconfiguring of the
syntax allows mántram to be the obj. of vanóṣi, with the point being that Agni, by
properly performing his ritual duties, acquires a powerful mantra for the weakling in
his charge. (The perhaps overly heavy tr. “mental spell” was meant to highlight the
etymological relationship with mánas- ‘mind’.)
My interpretation differs from the standard in other smaller ways. In c I take
dhā́yase not as a quasi-infinitive to √dhā ‘place, establish’ (e.g., Ge “um sich
Sicherheit zu schaffen”), a formation not otherwise found, but to the standard s-stem
dhā́yas- ‘nourishment’ to √dhā(y) ‘nourish’. I also take kīréś cid with the preceding
dative phrase, rather than with mántram.
I.31.14: Again my interpr. differs from the standard, though not as radically as in vs.
13. Most tr. take ab as a single clause with, implicitly, an embedded nominal relative
clause spā́rhaṃ yád réknaḥ paramám. The main clause resumes with vanóṣi, hence
the accent, and the yád clause is picked up by tád. The problem is that the RV does
not ordinarily allow embedded relatives; all relative clauses are either pre- or
postposed, as Hale has argued in the past. I therefore supply the frame “you are
35
kindled as protector” + DATIVE from the parallel vs. 13ab. Note that Agni as
protector also begins the next verse (15).
The phrase prá pā́kaṃ śā́ssi prá díśaḥ in d causes problems, because, despite
the strong superficial parallelism, the two accusatives pā́kam and díśaḥ do not appear
to be parallel. I have translated them as parallel, but admit the awkwardness. If the
verb prá śāssi is used in the same way in both iterations, the “quarters,” that is, the
geographical directions, are receiving the same instruction as the simple man. In
support of this interpretation ET remarks “my guess is that the poet intends a contrast
between two extremes (the limited simple man and the vast semi-divine quarters) as
recipients of instruction from the one who is superior to all in knowledge.” Ge supplies a
different verb with the second prá phrase: “du (gibst) Weisungen,” suggesting that
while pā́kam refers to the person who receives the instruction, díśaḥ might rather
express the content (“directions”) of the instruction.
I.31.15: In upamā́ I prefer to see the first attestation of the root noun compound
upamā́- ‘likeness’ (< ‘measure’), rather than the adverbial instrumental preferred by
Old and apparently followed by the standard tr. How an instr. would function here is
not clear to me, and Ge’s “der kommt zu oberst im Himmel” seems to push sópamā́
diváḥ further than the meager expression will take it. Old’s objection is that for such
a root noun we would expect nom. sg. upamā́ḥ, but this isn’t apposite: in this sandhi
situation upamā́ would be the outcome in any case, so it is only the Pp. reading that
enforces an underlying –ā́ final. Moreover, the parallel compounds pramā́- and
pratimā́- are attested in the RV with clear asigmatic nom. sg. (X.130.3), suggesting
that they have been reinterpreted as -ā-stems (see Scar p. 378). There is no reason
that an upamā́- wouldn’t have been treated similarly. See also upamā́ in VIII.60.13.
I.31.16: I interpret mīmṛṣaḥ slightly differently with its two different objects, as
“make X forgotten” and “make us forget X” respectively. The verb ‘forget’ is an I/T
verb of perception (in the terms of my 1983 book), and its causative thus can take
two different types of constructions.
Agni is “whirling” (bhṛ́mi-) because of the movements of his smoke and
flames.
[I.32 JPB]
I.33 Indra
I.33.1: There is no overt interrogative marker in b, but the kuvíd of c may suggest a
similar question in b.
The transitive thematic subjunctive vāvṛdhāti is assigned to the “Aorist des
Caus.” by Gr, but the properly formed redupl. caus. aorist ávīvṛdha- occupies that
slot. Other than our form, all forms belonging to the vāvṛdha- stem are medial, and it
seems best to follow Kü (471) in seeing this thematic stem as built to the perfect in
order to allow the root-final consonant to be maintained in forms like 2nd sg. impv.
36
vāvṛdhásva (since athem. *vāvṛtsvá could be taken to the root √vṛt ‘turn’). These
medial intrans./reflex. forms in turn generated the oppositional trans. act. vāvṛdhāti.
I take kétam páram in d as referring to Indra’s “distant intention” (which we
hope to move closer to us; see úpa in 1a, 2a), rather than our “highest wish,” as is
standard. The middle voice of āvárjate may support this. However the other
interpretation is certainly possible.
I.33.2: Given the importance of the close/distant theme in these verses and the
repeated úpa’s of 1a and 2a, upamébhiḥ in c should probably be rendered not only as
“best” but also “nearest/most intimate.”
I.33.3: Although the gen. in the rel. cl. yásya váṣṭi appears to be parallel to aryáḥ, as
I’ve tr. it, it may be better as a datival gen. “for whomever he wishes,” that is, Indra
redistributes cows belonging to enemies to his friends.
I’ve tr. the part. coṣkūyámāṇaḥ as an impv. to avoid the awkwardness of
“continuing to poke…”
In my opinion asmád ádhi continues the “distance” theme, and is equivalent
to the common āré asmát.
I.33.4: vádhīḥ … ghanéna, with verbal √vadh and nominal √han reverses the
expected distribution found, e.g., in VII.104.16 … hantu mahatā́ vadhéna (cf. also
I.94.9), with verbal √han and nominal √vadh.
The phrase ékaś cáran is a first instantiation of the lexeme famous in much
later times from the Rhinoceros(-horn) Sūtra. For the phrase in the dharma lit. and
the association with the rhinoceros (not, in my opinion, its horn), see Jamison 1998.
upaśāká- is only here, but śāká- is several times used of the Maruts or
Aṅgirases in their roles as helpers of Indra (IV.17.11, V.30.10). Therefore, though
there is a disjunction between “going it alone” and being accompanied by a host of
helpers, I take upaśākébhiḥ as personal here. The upa- may mark them as particularly
subsidiary, or it may simply have been prefixed to the stem because it is a signature
word in this part of the hymn.
Other tr. take sanaká- as the name of a group, but it seems a perfectly wellformed -ka-suffixed form of sána- ‘old’, with a pejorative diminuitive sense
appropriate to belittling one’s enemies. So Edgerton (1911 [k-suffixes]): 53): “the old
rascals.” My “old codgers” is also an attempt to capture the slangy and deprecatory
tone. On the demotic value of –ka- see Jamison 2009 (IIJ 52).
Ge takes prétim īyuḥ as an idiom “gingen in den Tod,” but despite the later
such usage of pra √i, this idiom does not appear in the RV. As indicated in the intro.,
I take this as referring to the separation of sacrificers from non-sacrificers.
I.33.5: Because the verb in d, adhamaḥ, is unaccented, something must be supplied to
complete the subordinate clause of c (prá yád …). It seems simplest to supply a form
of the verb √i ‘go’, esp. as prá √i is found in 4d. Ge’s solution, to supply the same
verb as in d, is possible, but seems pleonastic.
37
I.33.6: The Navagvas are ordinarily adherents of Indra’s. In order to preserve this
alliance, we must assume that the plural reference in pādas a-c alternates, with a and
c referring to Indra’s enemies, and b to his friends.
vṛṣāyúdho ná vádhrayaḥ is the clearest echo in this hymn of the phraseology
of the immediately preceding, very famous hymn depicting the Indra-Vṛtra battle:
I.32.7c vṛ́ṣṇo vádhriḥ pratimā́nam búbhūṣan “a steer who tried to be the measure of a
bull.”
citáyantaḥ is unclear, as forms built to this stem often are, and the tr. differ
appreciably: Ge “zu Einsicht kommend,” WG “erkennend,” Re “se rendant compte.”
In my opinion, it is in its usual intrans. value “appearing” and adds a simile-like
aspect to the main verb āyan, a sort of utprekṣā.
I.33.8: The verse is full of adornment/clothing terms; clear are śúmbhamāna‘adorning themselves’ in b and pari √dhā ‘clothe’ in c. In pāda a the middle
participle cakrāṇāsa- has a clear parallel in the adornment phrase in VIII.14.5
cakrāṇá opaśáṃ diví “creating for himself a headdress in heaven.” As in the previous
verses, there seem to be two contending sides, the enemies found in pādas abc and
the friends in d. “Having made for themselves a girdle from the earth” in a is easily
interpretable in this framework: the enemies have fallen and are perhaps dead. But
“adorning themselves with a golden amulet” in b is more difficult, since a golden
amulet sounds like a positive decorative item. However, Younger Avestan has a
compound zarənu-maini (Yt. 14.33), apparently made of related verbal material,
which is the epithet of a vulture, found in a passage in which the vulture espies
bloody meat from far distances. If there is a connection between the two (see EWA
s.v. maṇí), “to adorn oneself with a golden amulet” may mean figuratively “to
become food for vultures.” In contrast to Indra’s doomed adversaries in ab, in d he
clothes his “spies” with the sun; sunlight is often a symbol of untroubled life, as in
the often repeated wish “to see the sun” (sū́ryaṃ dṛśé and related expressions).
I.33.9: ET points out that the verb of ab pári … ábubhojīḥ “you coiled around” might
be more appropriate as a description of Vṛtra; it is almost as if Indra is appropriating
the qualities of his opponents in addition to his own and thereby showing himself to
be even more powerful. The form ábubhojīḥ itself is isolated, the only reduplicated
form to the root √bhuj ‘bend’. Kü gives it a lemma in his monograph on the perfect
(351–52) but does not commit himself further, beyond stating that the form is a 2nd sg.
preterite “in der statisch-attingenten Bedeutung” and suggesting that there might
have originally been a stative perfect that is no longer alive in Vedic. This would
allow a plausible analysis of our form as a pluperfect, with a renewed ending –īs, to
avoid expected but non-transparent *abubhok. JL points to the mirror-image phonetic
echo ubhe ábubhojīr across the pāda boundary.
adhamaḥ in d needs to be read with both pādas, c and d.
38
I.33.10: The rel. clauses with plural subj. in ab have no possible connection with
anything in the second hemistich. I take them instead as completing the portrayal of
the conflict between the two moieties depicted in the earlier parts of the hymn. The
pl. rel. prn. yé has no direct antecedent in the preceeding pāda (9d), though it can
pick up the intent of sg. dásyum; it can also hark a little further back to 9c
ámanyamānān ‘heedless ones,’ as H-P Schmidt suggests (B+I).
If pādas ab close the preceding myth, the rest of the verse seems to allude
glancingly to the Vṛtra (c) and Vala (d) myths.
I.33.11: abhí dyū́n: acc. pl. dyū́n almost always refers to ‘days’ (as in the expression
ánu dyū́n “through the days,” which regularly occupies this same metrical position.
However, pace Re’s “pour toujours” (which is, in any event, not equivalent to
“through the days”), a temporal interpretation does not work here. Ge (/WG) “für die
Himmels(götter),” for which there is no support (their I.190.4 is better tr. otherwise)
and whose datival “für” is an odd rendering of abhí. My “to high heavens” (the “high”
being imported from the English idiom) rests on the adj. abhídyu- ‘heaven-bound’; as
ET suggests it can be taken as a decomposition of this adjective, which, as it happens,
is almost always pāda final.
I.33.12: Bloomfield’s disc. of the 2nd hemistich (RR) is interpretively useful, though
somewhat dismissively phrased. It is too long to paraphrase here, but he acutely
observes that previous translators have glossed over the problem that Indra is
uncharacteristically depicted as at the end of his strength.
I.33.13: Stylistically the verse is marked by 4 fronted preverbs in tmesis, an effect
not possible to convey in English without awkwardness.
I.33.15: śáma- ‘hornless’ found only here and in I.32.15, another piece of shared
terminology.
I.34 Aśvins
I.34.1: “Three times a day” (tríḥ … adyá) opens the hymn, announcing the hymn’s
“three” theme and also linking it to the three pressings of the Soma Sacrifice.
predicative voc. navedasā here rendered as part of a phrasal verb, with impv.
bhavatam.
The second hemistich is built on an etymological relationship between the
instrument noun yantrá- (c) and the gerundive abhyāyaṃsénya- (d), both built to the
root √yam ‘hold’ (the second form presumably to the s-aor. of that root; so AiG
II.2.503, though the reason for using this base isn’t clear). This word play is lost in
Ge’s (/WG’s) tr., but I have aimed to keep it, unfortunately producing some
awkwardness in the English.
39
I.34.4: suprāvyè: prāvī́- and related forms we take, following Old (Noten ad II.13.9;
also Scar. 501) to √vī ‘pursue’, rather than to √av ‘help’ (e.g., Gr). The forms are
specialized for the pursuit of ritual activities.
tredhéva “as if threefold” presumably refers to the Aśvins, who, though only
being a pair, are as effective as if they were three.
With Ge (/WG) I take akṣárā (in akṣáreva) as nom. sg. fem., corresponding to
the Aśvins, who are subject to pinvatam. See Old’s somewhat inconclusive disc.
(ZDMG 63 [=KlSch p. 310]) of the various options. Rivelex (I.16–17) takes it as a
neut. nom. pl. collective; Gr. also as a neut. pl., though without specifying case.
I.34.5: sū́re duhitā́ “daughter of the Sun” preserves, by most accounts, the archaic
sandhi of final –as > -e before initial dental. For further disc., see Jamison 2010 (Fs.
Melchert). The myth on which this pāda is based, Sūryā’s marriage, is not otherwise
mentioned in this hymn. On the formulaic representation of the myth in the RV see
Jamison 2001 (Fs. Parpola).
I.34.6: Ge (/WG) take mámakāya as referring to the poet himself (“… meiner
Wenigkeit”), with sūnáve in apposition and identifying the poet as the Aśvins’ son
(“als (eurem) Sohne”). This is certainly possible and in line with the self-deprecatory
use of mámaka- in I.31.11 for the poet-sacrificer’s self-reference. It is by no means
necessary, however, and it does introduce extra machinery.
I.34.7 ātméva vā́taḥ is an underlying grammatical pun. Though vā́ta- is of course an
Indo-Iranian word for ‘wind’, historically it is actually a thematized present
participle to the ‘blow’ root. Here I think it is an adjective ‘blowing’ qualifying ā̛tmā́,
which is itself being compared to wind in this simile. ātmán- here shows its older
‘(life-)breath’ sense, not the ‘self, soul’ that already begins to take over in the RV.
I.34.8: I take the kṛtám in b as predicated of both āhāvā́ḥ and havíḥ, with number
agreement with the nearer noun and slightly different senses (at least in English).
I.34.10: It is a physical oddity that the two Aśvins are apparently credited with plural
mouths. There are two possible solutions: 1) The plural expression has been adopted
from elsewhere. There is one other occurrence of madhupébhir āsábhiḥ (IV.45.3),
unfortunately also in an Aśvins hymn; however, other examples of āsábhiḥ occur in
plural context. 2) The mouths don’t belong exclusively to the Aśvins, but to other
soma-drinking gods. The 33 gods who accompany the Aśvins here for drinking in the
next verse might support this possibility.
I.34.12: The ca of d has no obvious function; Klein (DGRV I.227-28) ascribes the
construction to “looser nexus.”
I.35 Savitar
40
I.35.4: I take kṛṣṇā́ rájāṃsi loosely as an accusative of extent. Others (Ge, Re) supply
a verb to govern this phrase (‘verbreitend’ and ‘pour traverser” respectively), while
still others (Macd., Falk 1988, WG) take it as a second acc. with dádhānaḥ, as
appositive to táviṣīm “assuming the dark realms as his power.” This latter solution is
possible grammatically and does not require additional material to be supplied, but I
am somewhat dubious that the dark realms constitute his power.
I.35.6: This verse of cosmic mystery decked out in numerology comes as a surprise
after the simple, descriptive beginning of this hymn. The syntax of c is ambiguous:
Ge (Re / WG) takes amṛ́tā as nom. pl., supplying “him” as acc. with ádhi √sthā:
“Alles Unsterbliche ruht (auf ihm)…” I follow Old, who takes it as acc. pl., citing
III.38.4 ā́ viśvárūpo amṛ́tāni tasthau “Having all forms, he mounted on the immortal
(things?).”
I.35.7: In some ways a responsive verse to the previous one(s): ví … akhyat (a)
parallels 5a ví … akhyan; ciketa (c) responds to cíketat in 6d; and the three heavens
of 6a are alluded to in the query in 7d katamā́ṃ dyā́m “to which heaven (of three or
more)?”
I.35.8: In b I read trī́ with both dhánva and yójanā (taking both as neut. pl.). The
position of the numeral favors taking it with dhánva, which could, however, be
singular; in favor of reading the numeral (also) with yójanā is X.86.20 dhánva ca yát
kr̥ntátram ca, káti svit tā́ ví yójanā “Wasteland and chasm -- how many yojanas (of
distance) are they away (from here).” (Note that dhánva is singular in this passage.)
I.35.10: On svávām̐ as nom. sg. of the -s-stem sv-ávas-, see AiG III.287. The Pp.
analyses it rather as svá-vān ‘possessing property’. Curiously Macdonell (Ved.
Reader, ad loc.) claims to be following the Pp., but tr. ‘aiding well’.
Most tr. take pratidoṣám as a temporal, “towards evening” or “every evening.”
I think rather that it’s spatial, construed naturally with ásthāt ‘took his place, stood’.
The god is facing west. (Cf. Peter Pan: “straight on ‘til morning,” presumably
meaning ‘east’) The same expression, also of Savitar, is found in VI.71.4, though it
does not clinch the interpretation.
I.35.11: Although it is possible to construe c directly with d, it seems best, following
Ge, to supply a verb of motion in c. The d pāda has a close parallel in I.114.10,
suggesting that it is independent.
Note the unusual duplication of the nominal referent in both rel. and main
clause: yé … pánthāḥ … / tébhiḥ … pathíbhiḥ.
I.36 Agni
I.36.1: purūṇā́m is generally construed as qualifying yahvám, but this requires taking
the latter as an implicit superlative (Ge “den Jüngsten unter vielen”), which it is not.
41
(JL points out that a derivative of the real superlative, yáviṣṭhya, appears twice in the
hymn, vs.. 6, 15.) Better to take it as parallel to viśā́m, though since víś- is fem., not
modifying it as Proferes (2007: 31) does.
I.36.10, 17: As often it is difficult to know when to tr. analyzable words as PNs
rather than literally. There is no particular reason that I explicitly allowed both
possibilities in vs. 10 and only the PN in 17.
I.36.13: añjíbhiḥ would have been better rendered as ‘ornaments’ than ‘unguents’,
and as ET suggests, these ornaments could be vocal.
I.36.14: In keeping with I.37.14, it might be best to change ‘favor’ to ‘friendship’ for
dúvaḥ in d.
I.36.18: Ge (/WG) takes dásyave sáhaḥ as a phrasal personal name “Dasyave Sahas,”
a personal name that would have to be neuter. Better to follow Old (SBE) by taking it
as a qualifier of Agni. Re also rejects the personal name interpr., but considers it a
pāda-final truncation of the instr. sáhasā. (Because sáhaḥ can be construed without
problem as the nom./acc. neut. it appears to be there is no reason to resort either to
Re’s truncation or to an archaic instr. sg. zero-grade ending *-h1 in the mode of Hale
[Fs. Melchert].) Old’s interpretation finds support in 19, where the second pāda
contains an entirely parallel phrase qualifying Agni, with neut. noun construed with
benefactive dative: jyótir jánāya “a light for the people.”
I.36.19: ukṣitá- is ppl. to both √vakṣ/ukṣ ‘grow’ and √ukṣ ‘sprinkle’; both meanings
are apt for Agni.
I.36.20: I read ná twice in b, both as the neg. with the dat. infinitive prátītaye and as a
simile marker, evoking the common phrase mṛgó ná bhīmáḥ (I.154.2, etc.) and its
variants.
I.37 Maruts
I.37.1-2: An “improper” relative construction, with masc. pl. yé in 2a picking up
śárdhaḥ ‘troop’ of 1a, which is grammatically neut. sg., though conceptually plural.
I.37.3: The Pp. reads pl. káśāḥ ‘whips’ here, and standard tr. follow this, making it
the subj. of vádān. But kaśā- is otherwise only sg. in the RV, even when plural
entities wield it. I therefore take it as sg. káśā and as the subj. of śṛṇve, with the
Maruts as unexpressed subj. of vádān. This also makes better sense of the positions
of both eṣām and yád: most tr. construe eṣām with hásteṣu, which means the
unaccented pronoun would begin a clause. And yád would be too far to the right in
its clause: we expect ya-forms to follow at most one constituent. (Of course, it might
42
be possible to interpret eṣām káśā hásteṣu “the whip in their hands” as a single
constituent, but this would be pushing it.)
I.37.5: It is possible to take prá śaṃsā as 1st sg. subjunctive, as Re. does, though there
seems no compelling reason not to continue with a 2nd sg. impv., following the 2nd pl.
gāyata in the previous pāda (4c).
Pāda b seems an incipient izafe construction, with an appositive introduced by
yád -- even though it is grammatically impeccable as a standard rel. cl. (allowing for
the attraction in gender to neut. yád of putative *yám, whose antecedent ághnyam is
masc.).
Ge unaccountably interprets the med. pf. vāvṛdhe as a 1st sg. (“Ich habe mich
… gestärkt”; so also WG), though he doesn’t read prá śaṃsā in a as 1st ps. Although
this is grammatically possible, context suggests that the Marut troop is the subj. of
this verb.
I.37.6: Although nouns not in the vocative case generally lose their accent in
vocatival phrases (type sūno sahasaḥ “o son of strength”), the conjoined genitives
diváś ca gmáś ca retain their standard accent though part of the voc. phrase headed
by dhūtayaḥ “o shakers” – presumably in part because the pāda needs to begin with
an accented word.
I.37.9: The syntactic structure of this verse is not clear, in great part because it
contains no finite verbs, and most tr. leave the structure undefined. In my interpr.
pāda a is a causal clause dependent on the main clause of pāda b, with yát … śávaḥ in
the relative cluase of pāda c coreferential with the váyaḥ of b. The point is that at the
point of birth, in a stable situation (a), the Maruts had the strength to leave their
mother’s womb on their own (nír √i is a lexeme specialized for birth contexts; cf.,
e.g., IV.18.2, V.78.9) (b), and that same strength remains with them (c).
I.37.10: The standard tr. separate pādas a and b and supply a verb with the former
(e.g., Ge “stimmen”) with no obvious source. I prefer to take the two pādas together
and take kā́ṣṭhāḥ as an unmarked simile: “their songs (like) race-course posts.” The
lexeme úd √tan then has the meaning ‘stretch upward, erect’; this is the only finite
form of the idiom in the RV (or, it seems, anywhere), which is mostly attested in the
frozen adjective uttāná- ‘stretching upward, stretching out’.
Why their knees are bent is a matter of speculation, but it probably refers to a
crouching position suitable for driving (cf. VIII.92.3).
I.37.11: Although most tr. take the “child of mist” to be the rain, its physical
description here (“long and wide”) makes better sense for a cloud.
ámṛdhra- normally means ‘not neglectful, not slighting’, but this doesn’t yield
much sense here. I therefore take it in the passive sense ‘not (to be) neglected /
slighted’; English “not negligible” provides a perfect idiomatic counterpart.
43
I.37.13: The mutual chatter of the Maruts on their journey refers of course to the
thunder.
I.37.15: This final verse is oddly disjointed. For one thing there must be a change of
person from 2nd (vaḥ) to 3rd (eṣām) in ab in reference to the Maruts; it is difficult not
to take these pl. pronouns as coreferential. I supply dúvaḥ in a, because this pāda is
structurally similar to 14b; however, this is not strictly necessary, and an
interpretation like Ge’s “Denn es gibt etwas für euch zum Schwelgen” is certainly
possible. As for pāda b, it is generally taken to mean “we are theirs,” and, again, this
is possible. But given its structural similarity to the immediately preceding pāda, I
interpret it in similar fashion, as suggesting that we have something to offer them.
The last pāda then expresses what our service to the Maruts should bring about for us.
I.38 Maruts
I.38.1: For kadhapriyaḥ see I.30.20. In this passage the connection of this voc. with
the interrogative is esp. clear, since the pāda begins kád dha “what indeed?” kád is
translated twice, for clarity.
I.38.2: The gen./abl. form of diváḥ and pṛthivyā́ḥ is somewhat surprising, but, with
Old, it is best to assume they depend on kvà. Although the ná separating them is
somewhat surprising, it is possible to take it as a real simile particle rather than a
bleached connective (Re’s “aussi bien que…”). Since the point of this tṛca is the
anxiety occasioned by the Maruts’ absence from our sacrifice, the poet worries that
the Maruts have disappeared to some other sacrificer on earth as definitively as if
they had gone off to heaven.
In the simile in c, “in a pasture” is supplied on the basis of the formula √RAN
gā́vo ná yávase (V.53.16, etc.). Note that the expected yávase shows up below in 5a,
in a slightly off-kilter simile. This might be taken as “poetic repair” (see Jamison
2006: Paris poetics), but simultaneously “de-repair,” in that it introduces an element
from one verbal complex into another, where it is unexpected.
I.38.5: See remarks ad vs. 2 on the simile here.
The “path of Yama” is of course the path to death (or after death, to Yama’s
world). The prohibitive mā́ of pāda a must have domain also over pāda c.
I.38.6: durháṇā and related forms are most likely Middle Indic developments of
*durhṛ́ṇā (etc.) ‘evil rage’ from √hṛ ‘be angry’. See EWA s. HAR I.
Ge (WG) take párāparā as representing pára+apara- ‘further and nearer,
earlier and later’, but Old’s interpr. (followed by Re) as an āmreḍita preposition
‘further and further, ever further’ is more appealing. As Re points out, the adv. párā
and related forms are characteristic of nírṛti-.
44
I.38.7: The standard interpr. of avāta- here is ‘windless’, but with Gr (see also Lub) I
take it to the homonymous stem ‘unextinguishable, unquenchable’ (√vā ‘extinguish’).
The point here is that even in a waterless place the Maruts can make rain: wind is
irrelevant, but water that doesn’t give out is crucial. Cf. avánīr avātā́ḥ “unquenchable
streams” in I.62.10.
I.38.8: “lightning bellows” – a mixed image, of a type not uncommon in Marut
hymns.
I.38.10: Ge (WG) takes the sádma phrase as a parallel subject to mā́nuṣāḥ (requiring
a sg. form of reja- to be supplied), but an acc. extent-of-space interpretation works
just as well, without needing an extra verb. So also Re.
The last three verses (7–9) describing the thunderstorm are all couched in the
present tense, so the augmented imperfect árejanta is somewhat surprising. Vs. 10
does begin a new tṛca, however.
I.38.11: ródhas- is a bulwark or fortification (√rudh ‘obstruct’), in this context
indicating the ‘banks’ of a river, which keep the waters within.
I.38.12: The change of person between pādas, 2nd pl. vaḥ in a, 3rd pl. eṣām in b, is
exactly the same as in I.37.15 and equally inexplicable. I have therefore failed to tr.
eṣām. It is possible, of course, that eṣām doesn’t refer to the Maruts, but to part of the
listed equipage, perhaps the chariots – hence “Let your wheel-rims be steady, and
(your) chariots and their horses” – but the parallel structure in the previous hymn
makes that unlikely.
I.38.13: I have tr. jarā́yai as ‘to awaken him’, but this is probably wrong, however
appealing in context. The noun jarā́- only means ‘old age’, and therefore some
version of Ge’s “dass er das Greisenalter (uns schenke)” is better. Its intent would
match the last pāda of the previous hymn, I.37.15c “in order (for us) to live a full
lifetime,” and the two hymns have much in common.
I.38.14: The first two pādas contain two punning verbs, whose double meanings
reinforce each other: mimīhí can belong to √mā ‘bellow’ and √mā ‘measure’
(generally assigned only to the latter and so tr.). In the first meaning it refers to the
sound of the song, in the second to its regulated production, that is, to its meter.
tatanaḥ can belong to √tan ‘thunder’ and √tan ‘stretch out’ (Gr assigns to the former,
but standard tr. reflect the latter). In the first meaning it again refers to the sound of
the song, in the second, again to its method of production – in this case, the
prolonging of a tone or note. The second meanings of both verbs clearly belong to
the technical vocabulary of singing (see the next pāda, 14c, as well), the former
connect the poet’s sounds to those of the Maruts’ thunderstorm (cf. mimāti ‘bellows’
in 8a).
45
I.38.15: The two occurrences of tveṣá-, 7a and 15b, both referring to the Maruts,
should have been harmonized in tr. (currently ‘dazzling’ and ‘glittering’
respectively). A regrettable if minor lapse.
I.39 Maruts
I.39.1: mā́na- seems to activate the same pun on the homonymous roots √mā as was
noted in the preceding hymn, I.38.14. The Maruts project both their bellowing and
the measure of their song. (Standard tr. only recognize the ‘measure’ sense.)
“Like a flame” – what does śocíḥ correspond to in the frame of the passage?
Ge (/WG) take it as parallel to “you” (=Maruts), as the agent of throwing, and supply
an object “(ihren Schein).” But since śocís- is a neuter, it can as easily be an
accusative and express the thrown object, and this seems to me the more likely
interpretation (perhaps influenced by the modern flame-thrower). In a Marut context
it could stand for the lightning they deploy in addition to the thunder represented by
mā́nam. For further on this image see disc. at vs. 10.
I.39.2: The two contrastive pādas of the first hemistich express offensive and
defensive procedures respectively.
I.39.3: Standard treatments (including Old) divide pāda a into a rel. cl. and main
clause (e.g., Ge “Was fest ist, stosset ihr um”), accounting for the accent on hathá by
its placement immediately after the rel. cl. However, this interpr. requires taking yát
sthirám as an embedded relative, preceded by the preverb+part. associated with the
main verb (párā ha). Since RV does not have embedded relatives, it is best to take
yád as the subordinator for the whole hemistich (with domain over vartáyatha as
well). This also makes the two pādas more parallel: the Ge’s tr. of b as also
consisting of rel. cl. main cl. (“was schwer ist, bringet …”) is impossible.
I.39.4: The opening of pāda c is identical to 2c. The rest of this hemistich presents a
few problems. The phrase tánā yujā́ is rendered variously. My translation is based on
the observation that in almost every single instance yujā́ follows an instr. in an
expression meaning “with X as yoke-mate” (X may either be animate [e.g., I.8.4
tváyā yujā́] or inanimate [e.g., X.83.3 tápasā yujā́ “with fervor as yokemate”]). In
this case, I take the root noun tán- to refer to the Marut’s entire lineage, in other
words their family heritage and their sibling connections to each other. The instr.
phrase sárvayā viśā́ “with your whole clan” in the next vs. (5d) may convey the same
meaning. Taking tánā as ‘lineage’ here also has the merit of allowing a semantic
connection with tánāya in 7a.
I separate the two pādas (so als Re), in great part because of the position of nū́
cid, which usually opens its clause (here after an extra-sentential voc. rúdrāsaḥ).
However, a tr. similar to Ge’s, “your might is never to be open to challenge,” would
also be possible.
46
I.39.5: The two other occurrences of durmáda- ‘badly drunk’ (I.32.6, VIII.2.12) are
both in martial context and seem to refer to warriors intoxicated on the frenzy of
battle. The other occurrences are quite negative, whereas here we must take the word
as positive or neutral in describing the Maruts, who are, to be sure, frequently
depicted as being almost out of control. I think this is the point of comparison.
I.39.8: The threatening ábhva- (< privative á- + √bhū; cf. WG “Unwesen”) that
besets us represents the Vedic fear of formlessness.
I.39.9: The signature word of this final tṛca is ásāmi-, Ge’s “vollkommen”
“complete.” It literally means “without a half, not halved,” and the insistence on this
unusual form merits a literal translation, in my view, although it is less fluent than
the paraphrase.
káṇvaṃ dadá “you gave Kaṇva” -- the grammar is clear, but the meaning
somewhat odd. What is presumably meant is the ancient poet Kaṇva, ancestral to the
current line.
I.39.10: This verse shows a type of ring-composition with vs. 1, in sense though not
vocabulary. The hymn begins (1b) with a shooting/throwing metaphor, śocír ná
mā́nam ásyatha “(when) you cast your measure/bellowing like a flame.” The last
pāda of the hymn (10d) makes the shooting image more pointed: íṣuṃ ná sṛjata
dvíṣam “you launch your enmity like an arrow.” The “shooting” verb √as of 1b is
replaced by more generic √sṛj ‘launch, release’, but the simile specifies an arrow,
rather than the potentially destructive but less specific ‘flame’. The responsion
between these two expressions may give added support to the interpretation of śocíḥ
as an accusative in vs. 1; see discussion there.
I.40 Brahmaṇaspati
I.40.3: The sense of the hapax paṅktírādhas- ‘whose gifts come in fives’ is
unexplained. It may of course be some ritual reference (and the range of available
explanations tends in that direction), but I suggest that it might be a reference to the
fingers and mean that gifts come by the handful, that is, abundantly.
I.40.6: The ca in c is subordinating, as the verbal accent shows (pratháryathā). See
Klein: DGRV I.240.
I.40.7: Although the standard tr. take antarvā́vat as referring to territory “in between”
(e.g., Ge “das dazwischen liegende (Land)”), I follow Schmidt’s (B+I, 102)
suggestion that it is a pleonastically marked variant of antárvant- ‘pregnant’, an
interpretation that works nicely with the following verse. Although my translation
implies that antarvā́vat modifies kṣáyam, this cannot be true because kṣáya- is masc.
My rendering is an abbreviated form of “made his dwelling place into (something)
pregnant” (cf. Schmidt “seinen Wohnsitz hat er zu etwas gemacht, das … schwanger
47
ist”). Where I differ from Schmidt (and some others) is in the interpr. of pastyā̀bhiḥ
in pāda c. Opinion is split over whether this stem (and pastyà-) means ‘stream’ or
‘dwelling place’, and Schmidt goes with the former. Although that meaning works
well in this passage -- Schmidt takes pāda d as meaning “pregnant (with streams),”
that is, well-watered -- on balance the ‘dwelling place’ interpretation fits more
contexts better.
I.40.8: Most tr. (including Schmidt) render pṛñcītá with the anodyne ‘increase’, but
úpa √pṛc is a sexual idiom (‘inseminate’ < ‘engorge’; see, e.g., VI.28.8). The accent
on this verb probably results from its adjacency to accented hánti (on which see, HO
and JSK – reff.).
Note the gapping out of compound in the contrastive phrase mahādhané …
árbhe “when there is a large stake or a small,” where the independent loc. árbhe is
functionally parallel to the 1st compound member mahā-.
I.41 Ādityas
I.41.4: The voc. ādityāsaḥ was omitted from the tr. “O Ādityas” should be added at
the end of the first line.
I.41.8: The first two pādas have elicited a fair amount of discussion and disagreement.
The questions are these: 1) What is the nuance of práti √vac and, in particular, what
is the function of the accusative construed with it? 2) Are ghnántam, śápantam, and
devayántam parallel accusatives, or is there a dependency relationship among them?
Ge (/WG) take devayántam as the object of ghnántam and śápantam (Ge “der den
Gottergebenen schlägt oder flucht”), while Re (EVP V) and Old consider the three
accusatives parallel, with the first two given as negative examples, the third one as a
positive one. Although I think the latter view is correct, I do not follow these scholars
in their assessment of the function of the accusative with práti √vac: both take it as
the topic spoken about (as opposed to Ge and WG). However, though práti is rare
with vac (one other occurrence in the RV, VIII.100.5, non-diagnostic because it is
not construed with an accusative there), when that preverb appears with other verbs
of speaking (√vad, brū, ah), the complement identifies the person spoken to, not the
contents of the speech. I therefore take the clause to mean that “I” will not bother to
answer back to a man who behaves badly, either physically (ghnántam) or verbally
(śápantam), though I would to a godly man. The parenthetical “as if” could be
omitted here, as ET points out. With devayántam freed from its potential as object of
the first two participles, they can instead take vaḥ ‘you’ as their object (though
admittedly how a mortal can “smite” the Ādityas is a bit unclear).
I.42 Pūṣan
I.42.1: Note the regularly contrasting preverbs sám and ví opening the first two pādas.
48
Though Re and, judging from his tr., Ge take sákṣvā to √sac ‘accompany’,
following Gr., according to Narten (265 and n. 834, and already so in BR, Whitney’s
Roots, etc.) it belongs with √sah ‘conquer’. That prá √sah is fairly common in the
RV (including nominal compounds) while prá √sac occurs only once [jpb cites Brick
2002 for this; what is it?] may support this root assignment. I am, however,
disturbed by the position of the preverb and its accompanying enclitic pronoun naḥ:
the collocation looks like the start of a new clause. I wonder if we don’t in fact have
two imperative clauses here, the second with a gapped ihi. Cf. VIII.17.9 índra préhi
purás tvám. My translation “go forth … to victory” reflects this possible
interpretation.
I.42.4: tápuṣi- is attested only twice elsewhere, both times modifying hetí-, hence the
supplied ‘weapon’ here.
I.42.7-9: vidaḥ is formally an injunctive and therefore functionally multivalent.
Hoffmann (1967: 263) is not certain that it has modal value, though most tr. (Ge, Re,
WG) take it as an impv. (e.g., Ge “schaff hierfür Rat!”). The other question is
whether the krátu- that Pūṣan is to find is his own or meant for us (e.g., Re “procure
(nous)…”). I have chosen to take the injunctive not as an impv. but a future-oriented
indicative and to interpret the krátu- as Pūṣan’s, not ours – the point being that Pūṣan
will find the resolve at our sacrifice to fulfill the requests we make of him in the
imperatives. However, the other possibilities sketched above are not excluded.
I.42.8: My “with” tr. of pāda b obscures what I think the grammatical structure is.
Pace Gr and Old, navajvārá- should not be a bahuvrīhi, on accentual grounds, but
rather a karmadhāraya ‘new suffering’. The question is what relationship its pāda
bears to the previous one. Re’s suppressed purpose clause (or so his supplied “qu’il
n’y ait” suggests) may be the best possibility syntactically. If we simply assume an
imperatival “let there not be …” (so Maurer, sim. WG “nicht (sei uns)…”), parallel
to pāda a, we should expect mā́, not ná as the negative. It may be, however, that
existential prohibitives (of the “let there not be” type) are blocked, because the root
√as ‘be’ does not build an aorist and also lacks injunctive present forms, although
functionally similar mā́ bhūt does occur. I have not found a discussion of this issue in
Hoffmann 1967, but I may have missed it.
I.42.9: Though most of the verbs represented in this catalogue of imperatives
normally take objects, the rhetorical point of this listing is the stark abruptness, and
the inclusion of an object (udáram ‘belly’) with the final verb lays particular stress
on this last desire expressed, to eat one’s fill.
I.43 Rudra and Soma
I.43.1: Tr. of vocéma repeated for clarity.
49
I.43.2-3: It is striking that Aditi and her two most illustrious sons, Mitra and Varuṇa,
are depicted as closely connected with Rudra. It is not entirely clear why, though
perhaps it is simply an attempt in this relentlessly upbeat hymn to associate Rudra,
who can be viewed ambivalently, with these powerful and positive figures.
I.43.4: jálāṣa- is “of unclear meaning” (so EWA s.v.) and shows non-Indo-Aryan
phonology. In the RV it occurs twice independently and twice in this compound
jálāṣabheṣaja- (and one of its independent occurrences is adjacent to bheṣajá-),
always in association with Rudra. The translation ‘healing’ (see also Kuiper, Aryans
25-26, 46) therefore makes contextual sense, despite the lack of etymological support.
I.43.5: A verse-length relative clause, which can be construed either with vs. 4 (so
Ge) or, by my preference, with vs. 6 (so also Re, WG). There is no strong evidence
either way, however.
I.43.6: Global ref. to human kind as “men and women” is vanishingly rare in the RV.
I know only one other potential case, the identical dative phrase in VIII.77.8, though
it has a more restricted sense in that passage.
I.43.8: Although juhuranta and related forms are assigned to the root √hṛ ‘be angry’
by Insler (1968: 219ff.), an assignment accepted by Mayrhofer (EWA s.v. HARI),
such a meaning simply doesn’t make sense in this passage, and so I take the form to
√hvṛ ‘go crookedly, go amiss’. The phonology is perfectly apt: with a zero-grade
having vocalic u and consonantal r before vowel, despite the metathesized zero-grade
hru found in some forms.
I.43.9: The construction of the various parts of this rhetorically ambitious verse is not
entirely clear, and various tr. make various choices. (In addition to the standard ones,
see Lüders [Ṛta: 231–32] and Hoffmann [1967: 260].) I take amṛ́tasya as modifying
te, rather than construing it with prajā́ḥ (as, e.g., Ge does). I consider the two
locatives, párasmin dhā́man ṛtásya and nā́bhā, to be parallel and to express the two
geographically opposed places where Soma will be searching: highest heaven and the
navel of the earth (supplying pṛthivyā́ḥ with nā́bhā as often). The larger meaning of
this verse is addressed in the intro.
I.44-49: For illuminating remarks on the rhetorical and grammatical connections
among these hymns, see Jesse Lundquist XXX.
I.44 Agni
I.44.1: On the locatival -ar in uṣarbúdh-, see Lundquist XXX.
50
I.44.2: sajū́s-, opening pāda c and here rendered ‘jointly’, is etymologically related to
júṣṭa- ‘enjoyable, delightful’, which opens the verse, and the poet clearly recognizes
the connection. A tr. “in joint enjoyment with…” seems too heavy, however.
I.44.3: Our interpretation of bhā́ṛjīka- we owe to Thieme (Unters. 40 n. 2; see EWA
s.v. ṛjīka-).
adhvaraśrī́-. Pace Gr, Old (SBE) Ge (/WG), I very much doubt that -śrī́- in
this compound (or others) has transitive value: ‘das Opferfest verschönend’. For
extensive disc. see Scar (545-46), who lays out a number of interpretive possibilities
but seems to lean towards the one I also favor. (So also Re.)
Related to the analysis of this compound is the interpretation of yajñā́nām
adhvara-. Ge (WG) take the former as dependent on the latter (clearest in WG “der
die Opferhandlungen der Opfer verschönt”), but I think it more likely that the two
nouns are parallel and depend on –śrī́-, one as an independent gen., one as 1st
compound member. So Re.; Scar adduces VIII.44.7 adhvarā́ṇām abhiśríyam, with an
independent gen. of adhvará-, which supports this analysis.
I.44.4: The initial word of this verse, śreṣṭham is the superlative associated with śrī́-,
the last word of the previous verse. A translation “glory … / Most glorious…” would
have captured this connection.
I.44.6: All standard tr. take suśáṃsa- in an active sense (e.g., Ge “der …Schönes
sagt”) with the singer the recipient of Agni’s good speech. Certainly when applied to
mortals, this has to be the sense, but when applied to gods I take the adj. in passive
sense ‘good to proclaim/laud’, with here the singer in the dat. (gṛṇaté) as the agent of
the praising.
I.44.7: The very common epithet viśvávedas-, used especially of Agni, has a potential
double sense in all its occurrences: ‘possesing all possessions’ and ‘possessing all
knowledge’. In fact, although the latter is generally favored in RV translations, the
former may be the more stable sense, in that its 2nd member védas- is widespread in
the meaning ‘possession, property’, but not found as a simplex in the meaning
‘knowledge’. The parallel formation jātavedas- (see above, vs. 4) has a similar
problem. Though generally rendered as ‘who has knowledge of the beings’ it could
as well mean ‘who has possession of the beings’. Since jātávedas- seems to have
become more opaque to its users than viśvávedas-, which does often (?) participate in
its context, we do not translate jātávedas-.
I.44.8: Pāda a, with the list of gods in the accusative, is an expansion of devā́n in 7d.
Though b begins with a god’s name in the acc. (agním), this is to be construed with
the following pādas.
51
I.44.9: As with vs. 4, the beginning of this verse connects with the final word of the
previous one: 8d … svadhvara ‘o you of good ceremony’ / 9a … adhvarā́ṇam ‘of the
ceremonies’.
svardṛ́ś- is a difficult word, with multiple interpretations. See Scar (pp. 23439) for discussion of the various possibilities, though his favored one (“das
Sonnenlicht sehend”) seems to us the less common, since the adj. generally modifies
gods. We generally take it as meaning ‘having the look/appearance of the sun’, that is
‘looking like the sun’, but occasionally as ‘having the sight of the sun’, that is,
‘seeing the sun’. Interpretations like Ge’s “deren Auge die Sonne ist” are unlikely
because dṛ́ś- is not ‘eye’. See Re (EVP XII.81). In this particular passage, it is
possible that ‘seeing the sun’ might also be appropriate, as anticipating the
description of Agni in the next vs. (10b) viśvádarśata- ‘visible to all’. In other words,
the gods in 9 “see the sun” and in 10 Agni, often homologized to the sun, is
something that everyone sees.
I.44.10, 12: In my opinion, purohitá-, though preserving its literal meaning ‘set in
front’, also already refers technically to a priestly office, the figure later known as
the Purohita. Agni is called purohitá- both because he is literally ‘set in front’, that is,
moved to the east to serve as the offering fire (later called the Āhavanīya), and
because he serves as priest. JPB, however, does not believe that the word has
developed this technical meaning in the RV. See esp. his 2004 “Bráhman, Brahmán,
and Sacrificer,” in Griffiths and Houben, eds., The Vedas: Texts, Language & Ritual:
Proceedings of the Third International Vedic Workshop, Leiden 2002, 325-44.
I.44.10: As just noted, this verse also shares lexicon with the previous one: 9d X-dṛ́ś-,
10b X-darśata-.
“Rich in radiance” is a less clumsy alternative for a literal rendering of the
bahuvrīhi vibhā́-vasu- ‘whose goods are radiance’.
I.44.11: Another lexical reminiscence across verses: 10d ..mā́nuṣaḥ ‘descendant of
Manu’ and 11c manuṣvát ‘like Manu’. As with the other examples, it is the last word
of the previous verse that is matched in the next.
I.44.12: Again, lexical echo, though in this case it’s the next-to-last word of the
preceding verse: 11d dūtám / 12b dūtyàm.
Note the synaesthesia in the second hemistich, with the sounds of the river’s
waves compared to the flashing of fire. The gen. agnéḥ here is a common noun
referring to the substance fire, not to the god. Its parallelism with síndhoḥ, also pādainitial, helps ground this usage.
I.44.13-14: The emphasis on hearing in these verses is continued in the next hymn
(vss. 2–3, 5, 7).
52
I.44.13: váhni- is usually ‘conveyor’, but here the gods cannot be conveying Agni,
and the word seems to be quite loosely connected with the notion of conveyance. My
‘passengers’ is probably pushing it beyond where it should go.
I.44.14: The final pāda makes a little ring with 2c.
I.45 Agni
The theme of “hearing,” found also in the last two verses of I.44, is further
explored here, with its complement, the “call” that the gods should hear.
I.45.1: My understanding of the structure of this verse is quite different from the
consensus, which takes yájā of 1c as a 2nd sg. imperative, addressed to Agni,
governing all the accusatives in the verse (“sacrifice to the Vasus…”). I instead take
yajā as 1st sg. subjunctive, governing only the accusatives of the 2nd hemistich, and
supply ā́ vaha ‘bring here’ from 2d to govern those in the 1st hemistich. (This is
supported by the fact that tráyastriṃśatam ‘three and thirty’ in 2d is a virtual
shorthand for 1ab vásūn … rudrām̐ ādityā́m̐ utá, the three divisions of the gods,
adding up to 33).
Despite the extra machinery, I think my interpretation better accounts for the
contrast between the accusatives in ab and cd: the first set names the large generic
groups of gods expected to attend the sacrifice, brought by Agni. The accusatives in
the second set do not fit this category; in fact, their most likely referent is Agni
himself: svadhvará- ‘of good ceremony’ is primarily and characteristically applied to
Agni, including in the previous hymn (I.44.8, a hymn that insistently associates Agni
with the adhvará- in 2b, 3d, 9a; see also 4c of this hymn), and mánujāta- ‘born of
Manu’ cannot be applied to other gods, but is appropriate to Agni; see mā́nuṣa‘descendant of Manu’ in the previous hymn (10d) and also used elsewhere of Agni,
as well as passages like VII.2.3 ... agním mánunā sámiddham “… Agni, kindled by
Manu.” It’s true that ghṛtaprúṣ- ‘ghee-sprinkling’ is not a particularly Agnian epithet,
though it could work if a passive interpretation of the root noun pruṣ- is allowed
(‘ghee-sprinkled’; cf. I.58.2), and that jána- is somewhat awkward as a designation
of Agni (see my uneasy ‘being’).
Still, the standard interpretation of the verse is more awkward: if the
accusatives in cd are held to refer to the gods (“the divine race”; see 10a daívyaṃ
jánam), they are described by adjectives that ill befit them; if Agni is held to be their
referent (as supported by the above arguments), then the verse calls upon Agni to
sacrifice to himself; if the accusatives refer to the human race, then the verse calls
upon Agni to sacrifice to humans, which is entirely contrary to the Vedic model of
sacrifice. By separating the two halves of the verse into different clauses and by
interpreting yájā as a 1st singular, I account for the different referent types of the two
sets of accusatives and avoid having Agni sacrifice to himself by providing another
agent for the verb in cd.
53
I.45.4: máhikeru- is a hapax of unclear meaning. Mayrhofer (EWA s.v. máhikeruand céru-) plausibly suggests a connection with √ci ‘observe, take note’; so also Old.
Its apparent structural similarity to mahivrata in 3c might invite a complementary
semantic analysis.
The etymological figure śukréṇa śocíṣā is not rendered so in English because
“blazing blaze” strikes me as limp.
I.45.8: Note the phonetic figure in c: br̥hád bhā́ bíbhrato havír, playing with b, bh,
and h; ṛ and r.
Most tr. take bṛhád bhā́ḥ ‘lofty light’ as coreferential with tvā (=Agni), not as
object of bíbhrataḥ, parallel with havíḥ. This in some ways makes better sense,
though the word order weakly favors my tr. Old (SBE) also takes the phrase as the
obj. of bíbhrataḥ and adduces a telling parallel, IV.5.1 kathā́ dāśemāgnáye bṛhád
bhā́ḥ, where the phrase is emphatically not coreferential with Agni.
I.45.10: I tr. sudānavaḥ as “you of good drops,” rather than “of good gifts,” which is
always also possible for this ambiguous stem, because of “the Maruts of good drops”
(marútaḥ sudā́navaḥ) who ended the previous hymn (44.14) in the same structural
position. But as a general descriptor of the gods in this verse it might be better as “of
good gifts.”
[I.46-47 JPB]
I.48 Dawn
I.48.1: The voc. uṣaḥ was carelessly omitted in the published tr.
I.48.2: With most tr./comm. I follow Bloomfield in interpr. viśvasuvíd- as haplology
for *viśva-vasu-víd-.
The pāda break favors taking bhū́ri with the verb, as most do (e.g., Ge “geben
sie sich viele Mühe”), but semantically it goes better with pāda a. Cf. expressions
like bhū́ri te vásu (I.81.2, 6, VIII.32.8), bhū́ri vā́mam (I.124.12, VI.71.4), and esp.
bhū́ri … saúbhagam in 9c below.
On sūnṛ́tā- as ‘liberality, liberal (gifts)’, see Re’s discussion here (EVP III.17),
summarizing previous work.
I.48.3: Most tr. take jīrā́ as agentive with an objective genitive (e.g., Ge “die Wagen
in Bewegung setzend”), but I think this unlikely because it would be the only such
usage of jīrá-. (gó-jīra- in IX.110.3 is sometimes so interpreted [Ge ‘die Kühe
zutreibend’] but need not be.) Although not taking jīrā́ as transitive leaves ráthānām
without any clear governing word, that seems preferable to claiming a unique value
for jīrá- in this passage. That Dawn may be “the lady of the chariots” is also
suggested by her hundred chariots in 7c.
54
The referent of yé in c and the affiliation of that rel. cl. are disputed. Ge takes
the rel. pronoun as referring to the ráthānām of b. This has the merit of associating
the rel. with an adjacent noun in the proper number and gender and keeping the
relative clause syntactically confined to the verse in which it appears. However, it
affords these chariots more agency and significance than I think they deserve. Instead,
the rel. cl. of cd seems to group more naturally with the identically structured yé
clause in 4ab, which also has a 3rd pl. presential reflexive verb preceded by a loc. pl.
and also contains a genitive referrring to Dawn. The two clauses also begin with a
metrically irregular 11-syllable pāda with a rest right before the caesura, an
irregularity that also speaks for their association. This pair of relative clauses are
resolved by the main clause in 4cd. Such an enjambed structure is characteristic of
pragāthas. (Re also take the two verses this way.)
Then there is the question of the meaning of dadhriré. It belongs to the root
√dhṛ ‘hold, support’, and in this (rare) middle usage seems to have reflexive value
‘hold oneself’/’hold oneself fast’, with the possible pregnant sense ‘hold oneself
ready’ (so Re, WG, Ge n.). I take it as having slightly different meanings in simile
and frame (a favored poetic strategy of Vedic bards). In the simile it depicts sailors
(or some sort of boatsmen) standing firm against the rigors of the voyage, whereas in
the frame the subjects (who are identified in the parallel rel. clause of 4ab as patrons)
hold themselves ready to give, an action that is also the topic of that parallel rel.
clause. In fact, one could almost construe (or supply) the dānā́ya of 4b with dadhriré
as well as with its own clause.
I.48.4: This verse is somewhat oddly constructed, especially the distribution of
elements in cd. The opening of c, átrā́ha tád, seems overburdened with functionless
elements, esp. the tád, which has no obvious referent. As it turns out, this opening is
found elsewhere (I.135.8, 154.6), with a likewise referent-less tád. I therefore assume
that the tád here emphasizes the temporal/logical átra. Then we find two gen. plurals,
eṣām and nṝṇā́m, separated from each other, but probably ultimately coreferential. I
assume that enclitic eṣām serves as the correlative for yé in pāda a (though we might
expect téṣām) and that the unusually heavy opening of the pāda has bumped it into
pseudo-second position after the first real word of the clause, káṇvaḥ. But until we
understand more about the interaction of the placement of these various elements,
this is simply an after-the-fact description. It should be noted that eṣām generally
does not show the standard Wackernagel’s Position behavior (modified 2nd position)
that we might expect from an enclitic, and in particular has a tendency to take final
position. The nṝṇā́m at the end of the verse simply doubles and further specifies eṣām.
I have tr. nā́ma twice, for ease of English.
: The first pāda depicts the usual effect of Dawn – sending all creatures on their
daily business.
ódatī: Though this form appears to be a fem. pres. participle to a Class I
present (also in its other occurrence VIII.69.2), such an analysis is formally
troublesome, because the feminine stem is weak (-at-ī), though a strong suffix is
55
expected in Class I (e.g., bhávantī-). Moreover, there are no other forms to the
putative present *ódati; the standard present is nasal-infix unátti with transitive value.
And ódatī- lacks participial sense: it simply means ‘wet’. It thus seems best to take it
as a non-participial –ant- adjective (as járant- is often interpreted). It is worth noting
that Whitney (Roots) classifies it as a primary derivative of the root and gives no
Class I present and that Gotō makes no mention of it in his monograph on Class I.
As for its meaning here, it is used simultaneously in two senses: the literal one,
‘wet’, referring to the dew characteristic of early morning, and ‘lubricious’, referring
to Dawn’s notorious hyperfeminine and sexual qualities, also reflected in 5ab yóṣeva
… prabhuñjatī́ “giving delight like a maiden.”
vājínīvant- (also vājínī-vasu-). This fairly common adjective is obviously a
derivative of extremely well-attested vājín- ‘prize-winning (horse)’, itself a
possessive adjective formed to vā́ja- ‘prize’. The usual tr.of vājínīvant- are rather
attenuated — Gr ‘gabenreich’, Ge (here) ‘du Reichbelohnende’, Debrunner (AiG II.2.
875) ’gabennreich’ — or render it as if it were identical to vā́ja-vant-; so Re. (here)
‘porteuse des prix de victoire’, WG ‘du Rennpreisbesitzende’. By contrast, we feel
that both the apparent feminine vājínī and the second possessive suffix (-vant- in
addition to –in-) should be noted and we interpret the stem as meaning ‘possessing
prize-winning mares’. The -vant-stem is esp. characteristic of Dawn and other female
figures (e.g., Sarasvatī), who might be expected to have female animals; though
vāj́inī-vasu- is almost entirely confined to the Aśvins, those gods are very closely
associated with Dawn. Debrunner (AiG II.2.409) instead ascribes the –ī- to
“Erweiterung durch -ī- nach Analogie anderer Wörter,” but doesn’t in this case
suggest what other word(s) might be involved.
I.48.9: The standard tr. take the injunctive uchat as a modal (Ge “soll … hinweg
leuchten”), but since this verb is parallel to a presential perfect nānāma (so Kü 27879, pace WG “hat sich … gebeut”) and a pres. indic. kṛṇoti, I see no reason to ascribe
modal value to uchat.
I.48.10: The publ. tr. “with your lofty chariot … heed our call” implies that the
chariot is the instrument of her hearing. This was not the intent: the chariot is simply
one of her attributes.
I.48.11: sukṛ́taḥ is multiply ambiguous. Though it literally means ‘doing/performing
well’ / ‘of good action’, it is ordinarily specialized for performing the sacrifice well
and refers to the human actors in the ritual. However, it can also on occasion be used
of gods (e.g., X.63.9). In this passage most tr. take it as a gen. sg. referring to the
sacrificer and dependent on adhvarā́n (e.g., Ge “zu den Opfern des Frommen”).
However, this leaves ā́ vaha without an object. I therefore read sukṛ́taḥ as acc. pl.,
and in fact I read it so twice — once as obj. of ā́ vaha and referring to gods (so also
Gr) and once as the goal of ā́ vaha and referring to the mortal sacrificers. The first
reading seems confirmed by the first pāda of the next verse, 12a, which “repairs” the
less clear expression with víśvān devām̐ ā́ vaha, using the same verb. The second
56
reading, referring to the sacrificers, allows the yé of 11d to have an antecedent of the
right grammatical number. It would of course be possible to take sukṛ́taḥ only once,
as referring to the gods, assuming the gapping of a pronominal antecedent to yé, but
this loses the neat equation of gods and sacrificers.
I.48.14: The relative clause of ab (lit. “which previous seers …” yé … ṛ́ṣayaḥ
pū́rve…) has no explicit referent in the main clause, but implicit is the notion that our
praises should receive the same favorable response from Dawn as theirs did, so there
is a suppressed gen. pl. téṣām or the like. The other passage containing the 3rd pl. mid.
pf. juhūré (VIII.8.6) is constructed almost identically to this one, though in dimeter
meter, with a similar implicit understanding of the relation between subordinate cl.
and main cl.
I.49 Dawn
On this hymn and, especially, on its “versified sandhi paradigm” (in JL’s
felicitous phrase) and the aberrant voc. uṣar (4c), see Lundquist XXX. The
appearance of this entirely anomalous form in the final verse of Praskaṇva’s Dawn
ritual series (I.44-49) defines this series as a type of ring: uṣar- as an –ar locatival
occurred in the first verse of the series (I.44.1) in the compound uṣar-budh- ‘waking
at dawn’.
The hymn itself has a simple ring structure: bhadrébhiḥ in 1a (with which
‘rays’ is supplied on the basis of I.48.13) matches ráśmibhiḥ in 4a, rocanā́t in 1b
matches rocanám in 4b, while the genitive diváḥ of 1b anticipated in 3d.
I.49.1: aruṇápsu- ‘of reddish breath’ is owing to Thieme (Fs. Schubring). See EWA
s.v. psu.
I.50 Sūrya
Although this is the last hymn in the Praskaṇva group (I.44-50), it does not
belong directly with the preceding hymns, which are clearly grouped in pairs and
belong to the Prātaranuvāka litany.
I.50.3: I separate pādas a and b and supply a verb of extension with b. Most tr. take
ádṛśram as the main verb of the whole verse. Although my interpretation requires
more machinery, it takes account of the fact that ví is vanishingly rare with the root
√dṛś (though it must be admitted that there’s an example in the nearby Praskaṇva
hymn I.46.11), while it is common with roots like √tan ‘stretch’, which also appears
regularly with raśmí-. Cf. X.129.5 vítato raśmíḥ. However, the standard tr. is
certainly possible. ET also suggests that since ví is fairly common with √bhrāj, that
might be the verb to understand with bc: “… his rays flashing widely through the
peoples like fires.” Of course, ordinarily participles with preverbs are univerbated,
but not always.
57
I.50.4: I don't understand Ge’s “pünktlich” for taráṇiḥ, which does not seem to
reflect any of the possible meanings of √tṛ ‘cross over, surpass, etc.’. Here the idea is
clearly that the Sun crosses the sky.
I.50.6: The first mention of any divine being but the Sun. The question here is
whether we have two additional gods or one. The two vocatives pāvaka and varuṇa
are in two different pādas (a, c), and the former is almost exclusively used elsewhere
of Agni. However, already in the hymn (vs. 1) another standard epithet of Agni,
jātávedas-, has been applied to a different god (Sūrya), and so it may be that aspects
of Agni, an alter ego of the Sun in some sense, are being distributed to other gods in
this hymn. Moreover, the sun is regularly considered Varuṇa’s eye, but not, I think,
Agni’s. Re, however, takes pāvaka as Agni.
The other question is the identity of the “bustling one” (bhuraṇyánt-). The
root √bhur and its derivatives are sometimes used of Agni and this referent is
possible here, but I think it more likely that it refers to the human ritualist, who is
active at the dawn sacrifice.
I.50.8–9: I assume that the seven mares of vs. 8 are the same as the sleek daughters
(śundhyúvaḥ … naptyàḥ) (or granddaughters) of the chariot, but I do not know why
the rare word naptī́- is used of them. Thieme (KlSch: 220), on the basis of
śundhyúvaḥ, thinks they are wild geese (“…hat sich sieben Wildgänse als Tochter
des Wagens angeschirrt”), but this seems to introduce an unnecessary complication
since śundhyú- can be taken in its literal meaning.
I.50.8: The reassignment of Agni epithets continues in this verse: śocíṣkeśa- ‘flamehaired’ is otherwise only of Agni.
I.50.9: sū́raḥ: Although most take this as nom. sg. (Gr, Ge, Th, Lü), Re makes a good
case for it as gen. sg. (followed, it seems, by WG), also adducing V.31.11 sū́raś cid
rátham.
I.50.12: ET comments “1.50.12 is AVS 1.22.4, AVP 1.28.4, the final verse of 4-verse
compositions for getting rid of jaundice. However, it's interesting that in both AV
recensions the first verse refers to the sun, but it's not the same as RV 1.50.11. The RV
seems to have a tṛca which incorporates the verse that appears as 1.50.12 whereas AV
tacks it on to a different tṛca. On the other hand, the 1st pl verb forms are jarring
in RV 1.50.12, but in the AV they harmonise with other 1st pls.” This formal and
structural argument suggests that the verse was taken over from Atharvan context, a
possibility that its contents also support.
I.51 Indra
I.51.2: This verse displays a type of “poetic repair”: the first hemistich has a
grammatical subject that is ordinarily inanimate (ūtáyaḥ ‘[forms of] help’) with a
verb that ought to have a personal subject (abhí .. avanvan ‘they attained to / gained’),
58
with an object possessing a number of qualities, but unnamed; the third pāda solves
this slight puzzle by giving names to both: the subject is the Ṛbhus, the object Indra.
The spatial contrast in b between filling the midpace, but being himself
enclosed by his own powers (muscle-bound?) is nice.
jávana- occurs only here in the RV, but the -ana-suffix ordinarily makes
transitive nominals (pace Ge’s “raschhandeln,” Re’s “véloce”).
I.51.3: Though the verse starts promisingly, with two identifiable myths (Vala, pāda
a; Atri, pāda b, though the 100-doored [house] is not otherwise known), the second
hemistich brings obscurity. As noted in the intro., the standard myth about Vimada
involves the Aśvins bringing him a wife, usually with the verb (ní) √vah. Is this the
same story, with vásu ‘good thing’ a generic substitution for ‘wife’, or is Indra’s
relationship with Vimada of a different sort than the Aśvins’? As for pāda d, the
action here is completely obscure (see Ge’s somewhat desperate note attempting to
make this about a rocky nest [Felsennest] of robbers), and what it has to do with the
Vimada story is equally obscure. Since nartáyan in d is only a participle, it should be
attached to the main clause in c rather than relating a separate myth. A final bit of
obscurity is saséna ‘with grain’, which opens c. The stem sasá- generally shows up
in enigmatic phrases referring, probably, to the ritual grass and/or the cereal ritual
oblations.
I.51.4: This verse, by contrast, clearly concerns the Vṛtra myth and is for the most
part unchallenging. It is worth noting that its first pāda is structured almost exactly
like 3a and begins and ends identically: tvám … (a)vṛṇor ápa. The verse also
contains an occurrence of vásu (in b), which unfortunately doesn’t shed any light on
the mysterious vásu in 3c. In fact 4b is the only part of this verse that is somewhat
unclear: the dā́numad vásu (‘drop-laden goods’, taking dā́nu to ‘drop’ with Gr and
Re, rather than ‘gift’ with Ge [/WG]) is of course the water confined in the mountain
by Vṛtra, which Indra releases. But why does Indra hold it fast (ádhārayaḥ) in the
mountain rather than releasing it as usual? The passage is similar to the Indra
ātmastuti X.49.9 aháṃ saptá sraváto dhārayaṃ vṛ́ṣā. Perhaps he gave the waters, as
it were, emotional support – but this doesn't sound like either the Rigveda or Indra.
At best we’re left with an attenuated meaning like “help out.” Or — a long shot —
this is an expression like I.103.7 sasántaṃ … abodhayó ‘him “you ‘awakened’ the
sleeping serpent,” where abodhayaḥ is meant to evoke its opposite, ‘put to sleep’.
See intro. to that hymn and Jamison 1982/83. In that case ‘hold fast’ would evoke ‘let
go’. However, the formulaic nexus between √budh ‘awake’ and √sas / svap ‘sleep’ is
very strong, whereas √dhṛ is not regularly paired with, say, forms of √sṛj ‘release’,
and so I advance this possibility only very tentatively. ET offers another intriguing
suggestion. She cites the well-known Old Persian PN Dāraya-vahu (corresponding
phonologically to Skt. *dhāráya- + vásu, and wonders “Could the poet be deliberately
using, perhaps even punning on, an inherited Indo-Iranian collocation of the verb *dhṛ
with object *vásu?
I.51.5: Note alliteration: … pipror … prā́rujaḥ púraḥ, prá …
59
I.51.5: JL cleverly suggests that the verse contains a word play on the PN of Indra’s
defeated opponent Arbuda: by characterizing him as ‘great’ (mahā́nt-), the poet
implicitly evokes the semantic opposite árbha-, arbhaká-, which resembles the PN
phonologically and would help regularize the non-IndoAryan –b- in arbudá. So,
“you trampled down Arbuda (the little one), though he was great.” As JL points out,
support for this interpretation comes from 13a ádadā árbhām mahaté …, kakṣī́vate
vṛcayā́m … “You gave little Vṛcayā to great Kakṣīvant.
I.51.7: The phrasing of pāda b is conceptually backwards, strikingly so. Ordinarily
Indra drinks the soma and is moved to be generous, whereas here his (latent)
generosity rouses itself in anticipation of the soma.
víśvāni carelessly omitted in publ. tr.: “all the bullish strengths.”
More alliteration: vṛścā́ śátror áva víśvāni vṛ́ṣnyā, with sequences of v with
either i or ṛ, followed by ś or ṣ (with a few more v’s and a ś thrown in).
I.51.9: The avratá- ‘having no commandment’ of 8b is transformed into the even less
savory ápavrata- ‘against/rejecting commandments’ and contrasted with their
opposite number, the ánuvrata- ‘following commandments’.
The image of Indra’s transformation into an ant (vamrá-) presumably
concerns his ability to pass unnoticed in the enemy camp and then bring the
fortifications down from within. However, “smashing apart” (ví √han) doesn't seem
a likely action for an ant, or even a huge nest of ants, so the combined image is
somewhat unsettled.
The identity of the enemy in this hemistich is not clear. The other occurrence
of the phrase dyā́m ínakṣant- (X.45.7) refers to Agni, but this identification seems
unlikely here. It should also be noted that the other genitive phrase referring to this
enemy, vṛddhásya cid várdhataḥ “the one who, though already full grown, kept
growing,” is grammatically problematic because the active present participle
várdhant- should be transitive, as the rest of this extremely well-attested active
inflection is. Gotō (1987: 291) notes the problem but has no explanation either.
Expected middle *várdhamānasya would of course not fit this metrical position, but
that is not enough for a Rigvedic poet to contravene grammar. However, the active
part. more nearly matches the paired ppl. phonologically: vṛddha… vardha…, and
this may have influenced the poet to use the active form.
I.51.10: A nice adjacency figure, nṛmaṇo manoyujaḥ.
Ge (/WG) supplies “with strength” with pū́ryamānam ‘being filled’, but Re’s
“with soma” (an alternative allowed by Ge in his n.) seems more likely on the basis
of other “fill” phrases involving Indra. Esp. apposite is V.34.2, adduced by Ge,
where Indra fills his belly with soma while Uśanā offers him a weapon, much as here.
Indra’s exhilaration in the immediately following verse here (11a) also supports the
soma interpretation.
60
I.51.11: vaṅkū́ vaṅkutárā is generally taken as characterizing the speed of the two
horses, and I agree that that is the general idea -- but think this meaning arises
indirectly. Ge takes it as ‘flying’, Re as ‘rapid’, Hoffmann (Inj. 221) ‘ever faster
moving’, WG ‘ever faster galloping’, but this basic meaning does not fit the root to
which it most likely belongs, √vañc ‘move crookedly’, or the other occurrences of
vaṅkú-, esp. I.114.4. I think the nuance here is the same one found in the deriv. adj.
vákva(n)- ‘billowing’ à ‘surging’. The non-linear movement of the root is here
concretized as a wave motion, with the attendant speed and power associated with
waves.
The c pāda presents some difficulties of construction, particularly the two
accusatives yayím and apáḥ, which do not match in number. Ge and Re supply
‘mounts’ (ádhi tiṣṭhati) from the end of pāda b (or perhaps ā́ … tiṣṭhasi from 12a)
and ‘chariot’ with yayím and begin a new clause with nír. So, “the powerful one
(mounted) the speeding (chariot); he released the water in a stream.” Alternatively
Ge suggests that c is a single clause, but that apáḥ is not an acc. pl., as is usual, but
the rare gen. sg. construed with srótasā, so “the powerful one released the speeding
(chariot) with the water’s stream(speed) [mit des Wassers Strom(schnelle)].” Neither
of these fussy solutions is appealing. With regard to the latter, nir apáḥ [acc.pl.] √sṛj
appears to be formulaic (cf. I.103.2, X.124.7, the only other examples of níḥ √sṛj that
I know of), and so a gen. sg. is unlikely; with regard to the former, it seems overly
elaborate to supply so much material in a pāda that can be read as a unity. I follow
Old in taking yayím as an epithet of (so Old) or, better, an appositive or qualifier to
the waters. Since áp- ‘water(s)’ is in essence a plurale tantum, a parallel singular
would not be surprising. For yayí/ī- qualifying waters, cf. X.78.7 síndhavo ná
yayíyaḥ “coursing like rivers,” adduced by Old (also X.92.5). My tr. “for coursing”
rather than “as coursing” or the like is a concession to English.
I.51.12: Another verse with tricky constructions. In the first pāda the loc. vṛṣapā́ṇesu
goes misleadingly easily into English (“you mount the chariot to…” like “the bus to
town”). Despite my tr. I think it more likely that vṛṣapā́ṇesu is functionally a loc.
absolute of the type “when bullish drinks (are available)” à “on the occasion of
bullish drinks / when there are bullish drinks.”
In pāda b most tr. (Gr, Ge, Re, WG) take prábhṛtā as representing –āḥ out of
sandhi – following the Pp., hence a nom. pl. m. past participle – but as Old points out,
this is very disruptive to the syntax. Better, with Old, to interpret it as a loc. sg. to the
-i-stem prábhṛti- ‘presentation’, a possibility suggested by Pischel (see Old) and
mentioned by Ge in his n.
Pāda c is standardly taken as preposed to d and the verb is tr. as indicative
(e.g., Ge “du … deine Freude hast,” Re “tu prends plaisir”), but cākánaḥ is
undeniably subjunctive; yáthā + subjunctive regularly builds purpose clauses, which
are regularly postposed. I therefore take pāda c with ab: the purpose of Indra’s
mounting of the chariot is the pleasure he will receive at the soma sacrifice.
In d all tr. take ślókam as ‘fame’, but the noun refers rather to a very
perceptible noise or call that signals some event. The event is often the sacrifice and
61
the ślóka-, the noise, is often issued by the pressing stones (e.g., I.113.3, 139.10,
III.53.10); the noise of the ślóka- is loud enough to reach to heaven (e.g., I.83.6,
190.4). This pāda contains this same notion of the ślóka-, the audible signal of the
sacrifice, going to heaven, but it seems also, oddly, to suggest that Indra follows it
there. Perhaps this refers to Indra’s departure to heaven at the end of the sacrifice, a
common theme.
I.51.13: Indra’s transformation into a human female is no more surprising than his
changing into an ant in vs. 9, and is better supported. See Ge’s note.
I.51.14: The standard tr. take pāda b as a nominal sentence (“the praise song is a
doorpost”), but the verb of pāda a, aśrāyi ‘has been fixed, propped’, fits b very nicely,
as Old argues. Ge suggests such an interpretation in his notes, without rendering it in
tr.
The poet Kakṣīvant mentioned in vs. 13 is associated with the Pajras, who are
mentioned a number of times in the hymns attributed to Kakṣīvant. ET points out that
pāda b probably contains a pun on the PN pajrá-, which literally means ‘sturdy,
steadfast’, a meaning which works well with the fixed doorpost.
I take prayantā́ in d as a periphrastic future, not a straight agent noun.
I.52 Indra
I.52.1: The verb mahayā can either be a 2nd sg. imperative (so Ge [/WG]) or a 1st sg.
subjunctive (so Re). In favor of the former interpretation is the parallel initial verse
of the last hymn, I.51.1. abhí tyám meṣám ... madatā, with imperative (2nd pl.); in
favor of the latter is the other main verb in this verse, 1st sg. opt. vavṛtyām. Either is
possible; I weakly favor the 1st ps. subjunctive.
Since subhū́- ‘of good essence’ is adjectival, a noun should be supplied as the
subject of pāda b (pace Ge, who simply tr. “Kräfte”). The likely solution is found in
vs. 4 subhvàḥ svā́ abhíṣtayaḥ “his own superior powers of good essence,” and I have
supplied abhíṣtayaḥ here. (So also, it seems, WG.)
The standard tr. take cd as a single clause, with the acc. índram of d identified
with the rátham of c. Although this is not impossible, turning the literal chariot of a
god towards the sacrifice is a common practice in the RV, just as turning the god
himself is, and an equation of Indra and the chariot is somewhat awkward. I therefore
think we have two separate clauses, with ā́ … vavṛtyām applicable to both.
The c pāda has, in my interpretation, a non-insistent but appealing syntactic
play, with the compound havana-syád- “rushing to the summons” parallel to the
simile átyaṃ ná vā́jam “like a steed (rushing to) the prize” — the suppressed term
being a form of the root √syand and the accusative vā́jam matching the first
compound member havana-.
I.52.3: A challenging verse, describing Indra in unusual ways and deploying unusual
words and constructions.
62
The first pāda contains the difficult but clearly related words dvaró dvaríṣu,
which seem also to belong with vṛ̥ḱ a-dvaras- (II.30.8) ‘having the X of a wolf’.
Wackernagel (1918 [see details in EWA s.v. dvará-] = KlSch 325-26) adduces the
Avestan root duuar, which expresses a daevic way of moving. If vṛ̥́ka-dvaras- means
‘having the movement/gait of a wolf’, I tr. the words in this passage as ‘skulking,
skulker’, as characteristic of a wolf.
The rest of the first pāda consists of vavrá ū́dhani. The latter is clearly a
locative, but the former is taken by the Pp. as vavráḥ, nom. sg. of vavrá- ‘cave,
cavity’ out of sandhi, an interpretation followed by the standard tr. and argued for by
Old. (Gr, however, takes it as a 3rd sg. pf. to √vṛ ‘cover’, vavré.) The sense is taken to
be “a cavity at the (soma) udder”; that is, Indra’s mouth, throat, and stomach are an
enormous empty space to be filled with soma. By contrast I take it as a loc. to the
same noun vavrá- and a simultaneous reference to the Vṛtra myth and the Vala myth,
as well as fitting the image conjured up by the dvar-words. To start with the last,
caves are good places to skulk and quite possibly a haunt of wolves. As for the Vṛtra
myth, Vṛtra himself is called a vavrá- in V.32.8, while Vala is itself a cave and the
word vavré is several times used of this myth and Indra’s involvement in it (IV.1.13,
V.31.3). Thus Indra is “skulking” in the vicinity of these mythological enemies in the
first part of this verse. The published tr. limits the reference of vavré to the Vṛtra
myth; I would now expand that.
I then take the adjacent loc. ū́dhani as contrastive and construe it with pāda b:
Indra skulks near his enemies (the “cavities”), but at the (soma-)udder he becomes
roused to elation and display his golden foundation, that is, the riches he will
dispense in return for the soma. Indra’s bright budhná- here contrasts with the
budhná associated with Vṛtra in vs. 6, where the latter lies on the budhná- “of the
dusky realm” (rásasaḥ).
The last part of the last pāda, sá hí páprir ándhasaḥ, is also problematic. It is
universally interpreted as “he is filled / fills himself with soma,” which makes good
sense. Unfortunately it does violence to the grammar. First, pápri- does not
otherwise mean ‘filling’ (in my opinion, but see, e.g., Grestenberger, JAOS 133.2:
271, though she does not give exx.), but either ‘providing’ or ‘delivering’.
Furthermore, reduplicated –i-nominals are otherwise agentive (AiG II.2.291-93) and
regularly take accusatives (see esp. VI.50.13 dā́nu pápriḥ ‘supplying gifts’)(see
Grestenberger JAOS 133.2). Ge is aware of the morphological problem (though not,
it seems, the semantic one) and in his n. suggests that the form is either reflexive or
that jaṭháram ‘belly’ should be supplied, but there is no basis for either of these
solutions. Therefore, although I see the attractions of “is filled with soma,” I do not
see a way to wrest this meaning out of the text. Instead I take ándhasaḥ as a causal
ablative and pápriḥ in the same fashion as VI.50.13. The clause then paraphrases
pāda b: Indra provides wealth because he becomes exhilarated on soma.
I.52.4: It is not clear to me why Indra’s superior powers have barhis as their heavenly
seat, but this does not license the grammatically impossible tr. of Ge and Re, who
seemingly take sádmabarhiṣaḥ as modifying índram.
63
For avātá- see comm. ad I.38.7 and VIII.79.7.
I.52.5: svávṛṣṭi- only here (and 14c below, in the same phrase), and the etymological
relation of vṛṣṭi- is not clear. Easiest (with Gr) would be to take the second member
as vṛṣṭí- ‘rain’, but ‘having his own rain’ doesn’t make much sense. Ge (n. to 14bc)
connects it with várṣman- ‘height’, várṣiṣṭha- ‘highest’, vṛ́ṣan- ‘bull’, tacitly positing
a root √vṛṣ ‘be high/great’ and tr. ‘Eigengrösse’. One of the difficulties with this
interpretation is that the word should be a bahuvrīhi (so Old) not a karmadhāraya,
judging from parallel formations (cf. svá-yukti, svá-vṛkti [pace Gr, Old]). I prefer the
interpretation that links the word to the IE root *u̯erǵ ‘work’, found in Aves. varəz
(and of course Engl. work, Grk. ἔργον)(see EWA s.v. svávṛṣṭi-). So, evidently, Re:
“son action propre,” though Re also takes it as a karmadhāraya. Because of the
formal parallels, I interpret it as a bahuvrīhi ‘having his own work’, even though this
causes some problems: in this clause Indra must be referred to both in accusative, in
this compound, and in the genitive, in the phrase asya yúdhyataḥ, which depends on
máde. Nonetheless, as usual I don’t feel we can ignore grammar whenever it
complicates interpretation.
I.52.6: durgṛ́bhiśvan- clearly belongs with durgṛ́bhi-, but the –śvan- is curious.
Probably best to explain it, with Scar (116) as a Kunstbildung based on ṛjíśvan- and
possibly mātaríśvan-. For this reason I’ve translated it as a nickname.
I.52.7: I take yújyam as having gerundive force, construed with te, rather than simply
‘his own’ < ‘associated (with himself)’ of other tr.
I.52.9: Another puzzling verse, and my interpretation is accordingly not at all certain.
I take the first pāda, couched in the neuter, to refer to the sun (n. svàr-), the placing
of which in heaven (as m. sū́ryam) was Indra’s last act in vs. 7. In the 2nd pāda the
subjects of ákṛṇvata (note the middle, which should have self-beneficial force), make
this sun into their own means of getting to heaven. As an -ana-nominal, róhana(only here) should have transitive-causative force.
However, I think there is more going on here, for in the 2nd hemistich Indra is
identified as the sun (n. svàr), while his helpers, the Maruts, are associated with
humans, the descendents of Manu (mā́nuṣa-), and their activities. If Indra is the sun,
then the sun of pāda a, which the Maruts/gods used to get themselves to heaven in
pāda b, may well be Indra. For this identification note the -(ś)candra- reminiscent of
Indra’s candra- in 3b, and in 6a the glowing heat surrounding Indra and his flaring
power seem to depict something very like a solar Indra. The Maruts’ aid to Indra in
the Vṛtra battle (4c, where they are called ūtáyaḥ as here) stood them in good stead,
enabling them to bridge the distance between the human world and heaven by
hitching their wagon to a star (=sun, =Indra).
I do not quite understand the bhíyasā of b, though it obviously must be
considered in connection with the same word in the same metrical position in the b-
64
pāda of the next verse. I assume it refers here to the awe- and fear-inspiring aspects
of Indra in his celestial form.
I.52.10: I agree with Ge (against Pp, Gr, Old, Re, WG) that loc. vájre should be read
for Pp. nom. vájraḥ and that this locative is functionally, but not grammatically,
parallel with áheḥ svanā́t “from the sound of the serpent.”
With Ge and Old (and back at least to Ludwig), I see no choice but to accent
the apparent voc. rodasī as ródasī. In the publ. tr. it should therefore be marked with
an asterisk.
I.52.11: I supply a form of √tan ‘extend’ in the first pāda, though with a general
injunctive sense, not the subjunctive of tatánanta in b. The “ten coils” of pāda a
invite an interpretation of increased or increasing space, as do the next verses with
their emphasis on distance and vast space.
I.52.13: The 2nd sg. act. forms bhuvaḥ and bhūḥ that serve as the main verbs of the
first two pādas respectively are difficult to distinguish. (Note that Hoffmann [Inj.
214–15] translates them both as “bist.”) The problem is made more acute by the fact
that though bhūḥ is definitely a root aor. injunctive, bhuvaḥ can either be an
injunctive to the marginal and secondary thematic aor. stem bhúva- or the
subjunctive to the root aor., as it is, in fact, in 11d. I have made an effort to
distinguish them in tr., and given the general preterital cast of this verse and the
previous one I am reluctant to interpret bhuvaḥ as subjunctive (“you will become the
counterpart of earth”), though that interpretation is not beyond possibility.
I.53 Indra
I.53.1: I am puzzled by Ge’s (/WG) interpretation of this pāda, which introduces a
thief with no support from the text (“Noch nie hat ja einer das Kleinod wie (ein Dieb)
bei Schlafenden gefunden”). As far as I can tell, the purport is that it’s easy for a
thief to find (and presumably steal) a treasure that belongs to people who are asleep,
but not so easy for us to do so in this case. WG remark that stealing something from
sleepers is a favored theme in later literature. But it is not otherwise met with in the
RV, as far as I know, and it doesn’t fit the context very well. I think the point is
rather simpler: we had better get to work presenting our praise to Indra because the
lazy and somnolent don’t get rewarded – “asleep at the switch” is an English idiom
for people who don’t pay attention.
I.53.2: The slightly slangy tone of the previous verse is continued here, in the
repeated verb duráḥ ‘break out’ and the cpd. ákāmakarśana- ‘not shorting desires’,
as well, perhaps, as śikṣānará- (for which see AiG II.1.315–16).
I.53.3: mā́ … kā́mam ūnayiḥ “don’t leave the desire lacking” matches the compound
akāmakarśanaḥ “who does not short their desires” in 2c.
65
I.53.6: tā́ni vṛ́ṣṇyā can be either nom. or acc. Most tr. opt for the former, but I do not
see how “bullish powers” can be the agent of exhilaration in the same way that soma
drinks are. Surely the point is to rouse Indra’s bullish powers for the fight to come.
Ge (/WG) take dáśa … sahásrāṇi as “ten thousand,” while Re separates the
two numbers as I do. The former interpretation is certainly possible, although the
distance between the words mildly supports taking them separately. The compound
numbers in vs. 9 are adjacent to each other. However, note navatím … náva ‘99’ in
I.54.6d.
A little phonological play: barhíṣmate … barháyaḥ.
I.53.7: Note the parallel complex double figures opening pādas a and b: yudhā́
yúdham and purā́ púram, with instr. and acc. sg. of a root noun in each instance.
I.54 Indra
I.54.1: The mā́ prohibitive lacks a verb, and there is nothing nearby to supply. The
universal solution, “leave, abandon,” does the trick, although it would be nice to have
some support for it.
róruvad vánā is variously interpreted. I have taken vánā as extent-of-space,
though construing it as a second object of ákrandayaḥ (WG) would also be possible,
save for the fact that the same phrase recurs in 5b and WG must construe it with a
different verb. There seems no reason to supply a separate verb to govern it, as Ge
does: “(du knackest),” and taking vánā as agreeing with róruvat as Re does (“les
arbres (ont) grincé-violemment”) introduces unnecessary grammatical complications.
(Is he thinking of this as a variant on neuter pl. + sg. verb?) For a expression similar
to my suggested interpretation see váne … vacasyate “display his eloquence in the
wood” in the next hymn (55.4).
I.54.3: The construction of the second hemistich is not entirely clear. Most tr. take
barháṇā kṛtáḥ together (e.g., Re “créé par une pression-violente”), but this requires
supplying a verb with the first part of pāda d (e.g., Re “(s’est mis)”). I instead think
the idiom is puráḥ √kṛ ‘put in front’ (I.102.9, VIII.45.9, X.171.4, of which the first
two have ‘chariot’ as obj. – e.g., VIII.45.9 rátham puráḥ … kṛṇotu). I do not take
háribhyām as an ablative, because 1) puráḥ + ABL is only dubiously attested, and 2)
setting Indra-as-chariot in front of his horses would be literally putting the cart
before the horse. I take háribhyām as dative, and think the idea is that Indra/the
chariot is set out front for the horses, that is, for them to be hitched up.
Ge and Re take vṛṣabháḥ with rátho hí ṣáḥ, but this is basically impossible,
given the position of the hí, which overwhelmingly takes 2nd position. Nonetheless I
agree that Indra is being identified with the chariot (not, however, with Ge the
chariot(-fighter)); WG supply “word” as the referent of sáḥ, but the striking equation
of Indra and chariot better fits the extravagance of the praise of Indra.
66
I.54.5: ní … vṛṇákṣi is here tr. ‘yank down’, whereas in the preceding hymn, 9d, I
render ní … avṛṇak as ‘wrenched down’. The two should have been harmonized in
the publ. tr. More serious is the question of what object the verb takes here. Most tr.
use vánā, which, admittedly, is the only available accusative, but I am reluctant to
follow this interpretation for two reasons: 1) As noted above róruvad vánā also
appears in 1c, which suggests that these words belong together and one shouldn’t be
extracted to serve as a complement for a different verb; 2) I really doubt that there’s
an alternate version of the Śuṣṇa story that involves felling trees on his head. In fact
Śuṣṇa himself serves several times as the object of (ni) √vṛj (I.101.2 śúṣṇam aśúṣaṃ
ny ā́vṛṇak, also VI.18.8, 26.3). In nearby I.51.11 Indra destroys Śūṣṇa’s fortified
strongholds (dṛṃhitā́ḥ … púraḥ), and I’m inclined to supply them here, with Indra
wrenching them down onto the head of their hapless defender. Note that Indra also
destroys púraḥ in the next vs. (6d). I would thus change the publ. tr. to “as you
wrench down (the fortresses) of the snorting Śuṣṇa onto his head.”
The question in the last pāda, kás tvā pári, lacks a verb, but it does contain the
preverb pári, which suggests the solution: pari √vṛj is a common idiom meaning
‘evade, avoid’, and since the root √vṛj supplies the main verb of the earlier part of
the verse (5a ní … vṛṇákṣi), there is support for supplying it here, with the pleasing
effect that the two different preverbs used with it provide two different idioms.
I.54.6: Support for supplying ‘help’ in pāda c (from āvitha in a) comes from
VIII.50.9 yáthā prā́va étaśam kṛ́tvye dháne, with the same root √av ‘help’ and the
same situation depicted.
I.54.7: As Ge notes, práti inoti is not otherwise attested, and so its sense here is
unclear (Ge “der sich an das Gebot hält,” Re “qui … va au-devant de l’ordonnance,”
WG “der … das Gebot entgegensendet”). I prefer to read the práti as adverbial ‘in
turn’, not a preverb, and ínvati in its usual transitive sense ‘drive, advance [smtg]’.
See I.55.4.
I.54.10: A poetically dense verse with striking images and concomittant difficulties.
The first problem is the isolated compound dharúṇa-hvara-, modifying támaḥ
‘darkness’ in pāda a. The compound is generally interpreted as a tatpuruṣa, with hvara- in verbal sense governing the first member (e.g., Ge "die den Urgrund der
Gewässer zu Fall brachte"), but the accent is wrong: we would expect final accent of
the type puraṃ-dará- ‘fortress-smashing’, brahma-kārá- ‘formulation-making’. By
accent the compound should be a bahuvrīhi (so WG “deren Wölbung ihr Grund
war”). The s-stem hváras- means ‘snare, tangle’ (from the meaning of the root √hvṛ
‘go crookedly’). I suggest that hvará- has a similar meaning and the whole
compound means ‘whose tangles were the foundation (of the waters: apā́m)’. And
what would this mean?
In order to decode it, we must first note the use of dharúṇa- elsewhere in the
Savya hymns: 52.2: párvato ná dharúneṣu ácyutaḥ “like a mountain, immovable on
its foundations” and 56.5-6: ví yát tiró dharúṇam ácyutam… “when you traversed the
67
immovable foundation” and … divó dharúṇam … pṛthivyā́ḥ… “the foundation of
heaven and of earth.” Given the connection of dharúṇa- with ácyuta- and párvataelsewhere, I think we can confidently take the támaḥ in a and párvataḥ in b as
coreferential (unlike Ge [/WG], Re). Remember also that Vṛtra is associated with
murky darkness (e.g., his lying “on the foundation of the dusky realm” in I.52.6). In
other words the mountain within Vṛtra’s belly in pāda b is the pure darkness of pāda
a. Its “tangles” represent the inability to see a clear path in the dark and may also
represent what happens to vision as it gets dark, the blurring and distortion of objects.
These tangles provide a foundation, and an enclosure, for the waters. If I am correct,
it is a powerful image.
The second hemistich is also problematic. At issue is the meaning of anuṣṭhā́ḥ,
which Indra smashes. The form must be acc. pl. feminine (though Scar [644] allows
the possibility of a nom. sg. masc., which would necessarily separate it from the
preceding víśvāḥ). The lexeme ánu √sthā straightforwardly means ‘stand by, stand
following, stand along’ and can be used for helpers who stand by a leader (as indeed
in nearby I.52.4); see exx. adduced by Scar. Scar then reasonably suggests that
anuṣṭhā́ḥ here refers to ‘Gefolgsleute’ (sim. WG). But this introduces a set of
subordinates and helpers to Vṛtra that do not otherwise figure in this well-known
myth. Ge takes it as “Einsperrer” (barriers), which makes sense but is hard to extract
from the form. Re’s “les positions-successives” is apparently an attempt to render
Ge’s translation in a lexically legitimate way, but it doesn’t make much sense. My
“rows (of palings) … in succession” is a similar attempt, with the palings a complete
invention. I do not feel that a satisfying solution has yet been reached.
I.55 Indra
I.55.1: phonetic figure … ví papratha, …pṛthivī́ … práti #
I.55.1–2: The two stems varimán- and várīman- appear here in successive verses
without clear differentiation in meaning (though they do appear in different
grammatical forms, nom. sg. and instr. pl. respectively).
I.55.2: The object of the verb práti gṛbhṇāti in the frame, which would correspond to
the rivers in the simile, is not expressed. Ge (/WG) supplies “die Somaströme,” Re
“chants.” Given the liquid nature of the simile, Ge’s suggestion seems the most likely.
Unfortunately most of the examples of ví √śri are used of the opening of the divine
doors in Āprī hymns, so there is no formulaic material to aid in determining what to
supply.
The phrase yudhmá ójasā is repeated in 5b and ójasā alone in 6b, both in the
same metrical position.
I.55.3: As Ge notes, √bhuj ‘enjoy, derive benefit’ is formulaically associated with
mountains, however odd that association may be to us. The question is then what
does Indra enjoy like a mountain. Ge takes it to be one of the elements in b, either the
68
‘principles’ (neut. pl. dhárman-) or the ‘manliness’ (neut. sg. nṛmṇá-), and interprets
masc. sg. tám in pāda as attraction from tā́ni or tád respectively. This is not
impossible, but I prefer to take the object in the frame as soma, which has the correct
gender and number, appeared in the previous vs. (2c), and is certainly something
Indra enjoys (although I have found no passages in which soma is explicitly
construed with √bhuj). The message of this first hemistich of vs. 3 – that Indra
displays manly power in order to enjoy the soma – is essentially the same as that of
2c, where he “acts the bull” to drink the soma.
Re rather trickily interprets the simile / frame construction with one verbal
expression in the frame (irajyasi “tu règnes sur”) and one in the simile (bhujé
“comme on jouit”), but this completely violates the structure of RVic similes, which
always hold the verbal notion constant between simile and frame. See Jamison 1982
(IIJ 24). WG supply soma, as I do, but also supply the verb ‘drink’ in pāda a and
separate it syntactically from pāda b. There seems no reason to do that.
In c I am very tempted to read devátāti with one accent, the loc. sg. of
devátāt-, rather than devátā + áti, with the adverbial instr. to devátā plus the preverb
áti. The meaning would be the same, and though prá √cit is very common, prá-áti
√cit would only occur here. For a parallel construction with prá cékite + instr. and
loc., see VI.61.13 prá yā́ mahimnā́ mahínāsu cékite “The one who by her greatness
shines ever more brightly among the great (rivers).”
I.55.4: What’s going on in this verse is a little baffling, but it seems to concern
Indra’s participation in the ritual as a (quasi-)priest-poet, speaking along with the
other priests (namasyúbhiḥ)(a) and (b) announcing his own name at that ritual. (That
‘name’ should be supplied here is clear from I.57.3, another Savya hymn, with nā́ma
indriyám.) Indra’s “singing along” with the human priests, as it were, is also found in
the passages adduced in Ge’s n. to 4a. It is a familiar topic.
Indra also seems to be homologized to soma in the first pāda: the only other
occurrence of vacasyate is found in a soma hymn (IX.99.6), where soma “displays
his eloquence” while sitting in the cups (camū́ṣu). Our word vána- ‘wood(en)’ is
often used in the soma maṇḍala for the wooden cup in which soma is put, and a wellattested formula combines váne, the bull (there =soma), and noisemaking, as here:
IX.7.3 vṛ́ṣā́va cakradad váne “the bull has roared down into the wood(en) cup” (cf.
IX.74.1, 88.2, 107.22). This superimposition of soma imagery on Indra contributes to
the obscurity of this pāda, esp. what “in the wood” means in reference to Indra. Ge (n.
4a) seems to think of a sort of summer camp in the woods for ṛṣis and their families,
while Re suggests a “marche” in the forest. I doubt both scenarios, although I do not
have a satisfactory solution of my own. If váne … vacasyate evokes the phrase
róruvad vánā of the immediately preceding hymn (54.1, 5), it can on the one hand
refer to Indra’s loud roar while doing battle in a natural setting; but in a ritual context
it might refer to the sacrificial posts or to the wood for the ritual fire, though I am not
entirely persuaded by either.
Indra’s benevolent aspect, despite his bullish nature, is emphasized in the
second hemistich.
69
I.55.5: As noted in the intro., this martial verse contrasts with the peaceful preceding
one, a contrast emphasized by their parallel structure.
A cute play in nighánighnate, where the preverb ni appears to repeat in the
middle of the word, although the second ni consists of the root-final n of the
intensive reduplication followed by an i-liaison.
I.55.6: This verse cannot be a single clause (as Re, WG seem to take it) because the
finite verb sṛjat in d lacks accent, while hí in the first pāda should induce accent on
the verb. But if we separate the last pāda from the rest, there is no main verb, just the
pres. participles vināśáyan and kṛṇván. Although present participles are rarely
predicated (as opposed to past participles), there are cases of such predication (pace
Lowe 2012), and I consider this one of them. In fact I connect the first three pāda of
this verse with the preceding verse, 5cd – with 6abc giving the reasons why the
people trust Indra – and the present participles in some sense reflect the intensive
(that is, iterative-frequentative) participle of 5d: he “is doing” rather than “did/does”
the actions; they are repetitive and ongoing.
I.55.7: “mind on” is the English idiom and is therefore used here, despite the Skt.
dative dānā́ya.
kéta- can belong either to gods or to men; here they must be Indra’s since they
are identified with his sā́rathi ‘coachmen, charioteers’. His intentions are presumably
to come to the sacrifice for praise and soma and, more to the point from our point of
view, to give to us, as expressed in the first pāda.
I.55.8: The etymological figure in b, áṣāḷhaṃ sáhas, rendered here with the
somewhat awkward “undominatable dominance” and belonging to the root √sah
‘vanquish, conquer’, is notable in part because the two root syllables ṣāḷh and sah
share no surface phonemes, since the past participle has undergone several regular
phonological processes that obscure its relationship to sah. Nonetheless any Vedic
speaker would instantly see the connection.
The simile in c is a little unclear in the absence of real-world knowledge of
life in Vedic India. WG suggest that, on departure from a temporary stopping place,
wells need to be covered over to avoid their getting filled in or otherwise damaged;
this seems reasonable, although I don’t see that this action needs to be restricted to
camps that are being left. In general it makes sense also in permanent settlements to
cover wells to avoid their being contaminated. In any case, the simile seems rather
more pointed and precise than necessary: that Indra has many hidden powers, mental
and physical, is a commonplace, and the image of wells seems, at least to me, a bit of
a distraction.
I.56 Indra
I.56.1: A bit of a mess, but very clever, once decoded.
70
For the first hemistich two features of interpretation are crucial: 1) I read
*avatásya ‘of the well’ instead of áva tásya, a reading already suggested by Gr (s.v.
áva). The ‘well’ word appears in the last verse of the preceding hymn (55.8c) and so
belongs to Savya’s diction. 2) The simile / frame structure of ab involves a
disharmony, with the verb to be interpreted in two different senses. In the frame, prá
… úd ayaṃsta, with the medial s-aorist to √yam ‘hold’, has a fairly literal meaning:
‘raised forth for himself’. The object is the “many dippers” (pū́rvīḥ … camríṣaḥ) of
the well (*avatásya). (In the publ. tr. “this” should probably be deleted, since tásya is
by my reading no longer there.) The word camríṣ- is found only here, but it appears
to be related esp. to camrīṣá- (I.100.12), apparently ‘beaker’, and the ‘cup’ words
(camasá-, camū́–) specialized for the serving of soma. The well is of soma; in
X.101.5-7 the preparation of soma is likened to raising water from a well. In the
simile (pāda b) the verb is used reflexively: the horse “raises himself up and forward”
to (mount) the mare, a pretty good representation of equine copulation. The acc.
yóṣām in the simile is not parallel to the acc. pū́rvīḥ … camríṣaḥ of the frame: the
latter is a direct object, while the former is a goal. The excitement of the mounting
stallion is implicitly transferred to Indra’s excitement at the many drinks of soma in
store for him.
In c I take dákṣam … hiraṇyáyam “golden skill” as a descriptor of soma:
golden because of its color, skill because drinking it gives Indra the ability to do
battle. It is a bit like calling alcoholic drinks “Dutch courage.” (In the next vs. Indra
is, or has become, the “lord of skill.”) The verb pāyayate is a lovely example of a
reflexive double I/T (in the terminology of my 1983 book): “he causes himself to
drink X,” with the appropriate middle voice. It is hard to know what (if anything) to
supply with mahé. I supply kárman- ‘deed’, whose only appearance in the RV is in
the preceding verse (55.3), Ge (/WG) “Kraft,” while Re takes mahé as standing for
an abstract, “pour (sa) grand(eur).” Any of these is possible; none is highly favored
over the others.
In d “ingenious” may not be the happiest tr. of ṛ́bhvas- as applied to an
inanimate thing. The word refers to craft or skill, and Re’s “habile(ment construit)”
may be the point.
I.56.2: On nemanníśaḥ see the lengthy treatment by Scar (55-56). I take párīṇasaḥ
somewhat loosely, following Gr, as an adverbial ablative.
In c Ge takes sáhaḥ as a pāda-final truncation of instr. sáhasā as sometimes
elsewhere, but this seems unnecessary. In the final verse of the preceding hymn
(55.8) Indra took sáhas- into his body. It does not seem odd that he would here be
identified as sáhas- itself. The odd placement of nū́ may support this analysis: the NP
vidáthasya … sáhaḥ may be structurally parallel to pa̛iṃ dákṣasya, and the 2ndposition nū́ could mark the second NP as a new syntactic unit.
I.56.3: “Like a mountain peak, … glints with its thrusting” – the image seems to that
of a pointed, snow-capped mountain, with the snow shining in the sun and the point
appearing to thrust into the sky, though of course it doesn’t move.
71
Again Ge suggests that pāda-final śávaḥ could be for instr. śávasā, though he
doesn’t so tr. – only wistfully remarks that ab could be a single sentence if śávaḥ
were instr.
I.56.4: arhariṣváṇiḥ is completely unclear; -sváni- is ‘sound, noise’, but the first
member appears nowhere else and has no etymology. All tr. take it as a cry of
triumph, but this unanimity reflects a dearth of other choices rather than conviction
in its rightness. Ge suggests arhari might be onomatopoetic, but it’s hard to see what
sound it’s imitating. One tack might be to play with MIA possibilities, but juggling
the phonology according to known MIA sound laws has not so far yielded anything
useful.
I.56.5: Though Gr and Lub take tiráḥ as the prepositional adverb, standard tr.
interpret it as the 2nd sg. injunctive to tiráti, which is surely the correct analysis.
There is a surprising disagreement among tr. as to what ácyutam ‘immovable’ should
modify. I take it with dharúṇam on the basis of Savya’s I.52.2 dharúṇeṣv ácyutaḥ. I
take rájaḥ as obj. of átiṣṭhipaḥ, despite the pāda boundary, since otherwise this
causative aorist is left without an expressed object.
I.56.6: pāṣyā̀ has neither an etymology nor a secure meaning; it occurs only once
elsewhere (IX.102.2), but since it is both times in the dual and in this case is used of
something belonging to Vṛtra that gets broken apart, “jaws” is a contextually
attractive translation. Savya’s I.52.6 vṛtrásya … nijaghántha hánvor indra tanyatúm
“when you, Indra, struck your thunder down upon the jaws of Vr̥tra” is similar.
I.57 Indra
I.57.1: Unlike the standard tr., I take d to mean not that his generosity is meant to
display his power, but rather that his generosity has opened up to, that is, has been set
in motion by his exercise of power.
I.57.2–4: A bit of word play in the sequence haryatá(ḥ) (2c), haríto (3d), harya tád
(4d).
I.57.2: Note the Wackernagel particle ha positioned between the preverb ánu and the
verb asat, despite the material preceding it in its clause.
With Ge I supply a verb of motion in b, because the “waters to the depths”
simile regularly appears with one (e.g., V.51.7 nimnáṃ ná yanti síndhavaḥ).
I.57.3: The phrase úṣo ná śubhre is quite problematic. In the first place, it is
syntactically odd to have a voc. in a simile (“X like o Dawn”). úṣaḥ may be vocative
by attraction from an underlying nominative, as in I.30.21 áśve ná citre aruṣi “O you,
dappled bright and ruddy like a(n o) mare.” Then, for reasons given in the intro., I
am certain that the fem. voc. śubhre in b cannot be addressed to the Sacrificer’s Wife,
72
despite the standard view, but that leaves the identity of the addressee baffling. Fem.
śubhrā́- is ordinarily used of Dawn herself, not someone or something like Dawn.
However, its other standard referent is Sarasvatī or another river or rivers (III.33.1-2,
VII.95.6, 96.2, V.42.12; waters V.41.12, maybe II.11.3; drops IX.63.26), so it is
barely possible that the water similes of vss. 1–2 here evoke an actual river to bring
the materials to the sacrifice. Better, but textually problematic: perhaps the identity
of simile and frame should be reversed, and the phrase means “O Dawn, like a lovely
(river), assemble …” (assuming an underlying *úṣaḥ śubhrā́ ná…). This would make
fine sense in the passage: Dawn comes at the beginning of the sacrifice, bringing
materials for it, and is compared to a river that picks up material from its banks.
Although this requires more manipulation of the text than I would like, a sequence
such as I just reconstructed, with the ná following two feminine singulars, might
have seemed anomalous and been restructure to a more conventional order: X ná X’.
On balance and with due caution, I endorse this solution and would now translate the
phrase as suggested above.
As for the object of sám … ā́ bharā, I supply ‘everything’, based on víśvam in
2a, also referring to the sacrificial materials.
The semantic basis for the simile in d is somewhat obscure. On the one hand,
the “tawny mares” (harít-) are often the horses of the sun, so that Indra has been
made a light (jyótiḥ) like the sun’s horses. On the other hand, Indra’s name and form
(dhā́ma … nā́mendriyám) are as suited for fame as horses are for running.
In d nā́yase is analyzed (starting with the Pp.) as ná áyase, but this produces a
bad cadence. I do not see any obstacle to assuming a preverb ā́, so ná ā́yase ‘for
coursing hither’, which fixes the cadence.
I.57.4: The translation “Here we are -- those of yours” reflects the annunciatory imé
as well as the te … té vayám, which identifies the speakers as Indra’s own.
In c I take cárāmasi as an independent verb, meaning to ‘carry on’ with life
and activities, though it is possible that it is an auxiliary verb with the gerund
ārábhya, as Ge takes it.
On kṣoṇī́- as ‘war-cry’, see Thieme (1978[79]: KZ 92: 46), EWA s.v.
I.57.5: táva smasi is a paraphrase of 4a (imé) te … té vayám
I.57.6: The “cut” in the first hemistich is ambiguous in English but is a past tense
rendering pf. cakartitha. I added ‘apart’ despite the absence of ví because unadorned
English “cut the mountain” sounds odd. The vájreṇa vajrin opening pāda b at least
provides the desired v-.
Verbal play, in which two unrelated words mimic an etymological
connection: párvatam (a) … parvaśáḥ (b) “mountain … joint by joint.” The two
items are in the same metrical position, and each is in a pāda that begins with an
alliterative pair: tuváṃ tám and vájreṇa vajrin.
I.58 Agni
73
I.58.1: Phonetic figure spanning the end of the first hemistich and the second (esp. its
end): ... vivásvataḥ# #ví ... , ... havíṣā vivāsati#
The lexeme ní √tud is generally taken to mean something like ‘spur on’, but
this tr. fails to render the ní. I prefer to take it in the literal sense ‘push down, force
down’, meaning that, in the English idiom, you can’t keep Agni down.
I connect pāda b with c, rather than with a, as the standard tr. do, because of
the difference in tense (pres. ní tundate, impf. ábhavat). This tense mismatch requires
the yád of b to have the sense “ever since” (Ge[/WG] “seitdem,” Re “depuis que”),
which does not seem to me to be natural to it. By contrast, Agni’s assuming the
office of messenger in b leads directly to his journey in c.
I.58.2: In b tiṣṭhati can mean, as I take it, “stays (within)” or, with Ge, “stands up
(in).” In the latter case, the image would be of a forest fire, fed by brush, flaring up.
This is possible, but in the absence of the preverb úd or similar directional
indications, I prefer the former.
I.58.3: On krāṇā́ ‘successfully’, derived from the old fem. instr. sg. of the med. root
aor. participle of √kṛ in adv. usage, see Old (Fs. Kern 33ff. [details in EWA s.v.
kṛāṇā́]).
Since ví + ṛṇóti/ṛṇváti regularly refers to the unclosing of doors (e.g., I.128.6
dvā́rā vy ṛ̀ṇvati), the tr. ‘distribute’ (Ge ‘teilt … aus’, WG ‘verteilt’, Re ‘répartit’)
seriously manipulates the idiom. I therefore prefer ‘disclose’ – that is, unclose and
reveal to sight. (So also Old SBE.)
I.58.4: The voc. rúśad-ūrme ‘o you possessing gleaming waves’ should, strictly
speaking, not be accented. It may owe its accent to IV.7.9 kr̥ṣṇáṃ te éma rúśataḥ
puró bhā́ḥ “Black is your course, (though) you are gleaming; your light is in front,”
with a gen. sg. part. rúśataḥ following an identical opening. (So tentatively Bl RR.)
Or perhaps as the first of two voc., in post-caesura position, it was felt to begin a new
syntagm.
I.58.5: On pā́jas- see Re ad loc. (n.; EVP XII) and EWA s.v., with lit. It seems to
refer to a surface or face, then to shape, area, or dimension in general, often with the
sense of “full dimension.”
I.58.5: On sthātúḥ (in the pair sthātúś carátham “the still and the moving”) as neut.
sg. to the -tar-stem (< *-tṛ) see AiG I.23, 301; III.204. Tichy (1995: 71) rejects this
explanation, but her alternative (a masc. nom. sg. to a –tu-stem) breaks the tight
rhetorical structure of this merism by pairing a masculine with a neuter (carátham).
Thus, whatever phonological problems there may be in assuming an *-ṛ# > -ur#
change (as opposed to *-ṛs > -ur as in the gen. sg. of -ṛ-stems), I favor the older view.
[I will not comment further on the phonological issue here, but JL suggests that there
might be a “poetic” derivation available in the formulaic material. Frotscher art.]
74
The nom./acc. neut form here would also be reinforced by the semantically identical
genitive sg. phrase sthāthúś ca ... (jágataś ca (I.159.3, II.31.5).
I.58.6: The last halves of the two hemistichs are parallel and complementary,
referring to humans and gods respectively: b: … jánebhyaḥ# / d: … divyā́ya
jánmane#
In c mitrám is of course ambiguous: it can refer simply to the god of that
name (so Re), but more likely is at least a pun on the meaning of the common noun
‘ally’. Agni is often so called because he serves as go-between between gods and
men.
I.58.7: The “seven tongues” (saptá juhvàḥ) are somewhat puzzling, or rather the
phrase has several possible interpretations. Re takes it as a “pré-bahuvrīhi” (probably
better expressed as “de-composed” bahuvrīhi), referring to the priests “having seven
offering ladles” (juhū́- meaning both ‘tongue’ and ‘ladle’); it is, of course, also
possible to take the ladles as subject without reference to an underlying bahuvrīhi (so
Old SBE, Ge), since inanimate things often have agency in the RV. But the other
meaning ‘tongue’ could also be meant literally (either in a de-composed bahuvrīhi or
not): (priests having) seven tongues, that is seven voices devoted to praising Agni.
See Ge’s n. (also WG). This interpretation would make the first pāda semantically
parallel with the second, where ‘cantors’ (vāghátaḥ) is the subject. And I will add
another, more distant possibility, but one that makes better sense of the ‘seven’ – viz.,
the seven rivers or streams. The seven rivers are credited with giving birth to and
nourishing Agni in a mystical passage in III.1.3–6, where they are also identified as
seven vā́ṇīḥ ‘voices’ (III.1.6d). Seven is a number especially characteristic of rivers,
and since rivers are often said to be noisy (indeed the word nadī́- ‘river’ is derived
from the root √nad ‘roar), calling the rivers “seven tongues” here would fit
semantically. In the end I don’t think that choosing one of these possibilities and
eliminating the others is the right strategy; the phrase is meant to be multivalent,
evoking a number of features of the ritual context.
I.58.8–9: Vs. 9 is essentially a paraphrase of vs. 8, with several parallel expressions.
And the final pāda of 9 is the Nodhas refrain.
I.59 Agni Vaiśvānara
I.59.2: On aratí- as a spoked wheel, which often serves as the symbol for the ritual
fire, see Thieme (Unters. 26ff., EWA s.v.).
I.59.4: The first pāda of this verse is metrically disturbed, which, in conjunction with
its syntactic incompleteness, leads some tr. to consider the text corrupt. I’m afraid I
find that reasoning too convenient.
This verse is variously interpreted, with its difficulties in great part arising
from the fact that there is no finite verb, but it can be decoded by paying attention to
75
the functional roles of the nominal complements. What seems to unify the verse is
the presence of a dative recipient in pādas a, cd, and I therefore (with most tr., but
not Ge) take the verse to be a single sentence, with the datives throughout referring
to Agni. The objects presented to Agni are songs (gíraḥ b), qualified as many (pūrvī́ḥ
c), and implicitly compared to exuberant maidens (yahvī́ḥ d), based on the fact that
gír- ‘song’ is feminine in gender. The grammatical subject is hótā in b, with parallel
subjects in similes: the two world halves (ródasī) in the first pāda and “skill”
(dákṣaḥ) in the second. The manuṣyàḥ in b I read twice: on the one hand, its position
directly before ná marks it as the first word of the simile, going with dákṣaḥ, hence
“Manu’s/manly skill,” but I believe it should also be read with the immediately
preceding hótā (“manly/human Hotar”). In Agni hymns Hotar is ordinarily
specialized as a designation of that god (see, e.g., immediately preceding I.58, vss.1,
3, 6–7), and manuṣyàḥ here makes it clear that the human priest is at issue, with Agni
himself the dative recipient. The last question is what verb to supply, and in a sense
the exact identity of the verb is not terribly important, as long as it has approximately
the right meaning and the right case frame. With Old (SBE, Noten) I supply ‘bring’
(√bhṛ), which is frequently used with gíraḥ and a dative recipient (e.g., I.79.10 …
agnáye / bhárasva … gíraḥ), but ‘sing, present, give,’ etc., would all work. I do not
see any reason for, or justification of, supplying a 1st-person subject, however, pace
Old SBE (1st pl.), Ge, Re (1st sg.).
I.59.5–6: 5d pāda here = VII.98.3d, of Indra, and Nodas uses a similar expression of
Indra in I.63.7d. Vs. 6 is even more Indraic. As noted in the intro., this part of the
hymn is designed to associate Indra and his great deeds with Agni.
I.59.7: The rest of pāda a essentially glosses vaiśvānaráḥ.
I take puruṇīthá- as a qualifier, not a personal name, contra most tr.
I.60 Agni
Taking off from my comment below on vs. 5, JL has further articulated the structure
of this hymn. What follows is mostly verbatim from his comments, with some
additions and light editing of my own:
I think this little hymn might have a slightly more elegant structure than has
been appreciated (I thought of this following your mention ad vs. 5 of the “faint
ring”). It seems to me that the 5 verses are nicely balanced rings within rings
revolving around vs. 3, the omphalos-like jā́yamānam… jījananta, harking back to
Agni’s double birth in 1c. The outer rings would be, as mentioned ad vs. 5, vss. 1c
rayím iva praśastám = 5 pátim ... rayīnā́m, prá śaṃsāmaḥ. Vss. 2 and 4 contain the
same word vikṣú; note esp. the alliterative and partly etymological figure in 2d
viśpátir vikṣú vedhā́ḥ. The hymn has not only a ring structure, but also forward
momentum provided by the movement from the larger social organization of the vís-,
in the full expression viśpátir vikṣú (2d), to the more intimate setting of the home,
emphatically presented in 4c dámūnā gṛ́hapatir dáme. The momentum can also be
tracked in the expressions of lordship involving páti-: viśpáti- (2d) to gṛhapáti- (4c)
76
and finally the solemnly pleonastic rayipátī rayīṇā́m (4d). Agni, celebrated (√śaṃs)
“like wealth” in 1c (rayím iva), is transformed into the lord of wealth in 4d. This final
title is repeated in 5a pátim agne rayīṇā́m, with his name interposed between the two
elements, and again the object of √śaṃs. The use of these three -páti- compounds
may convey the message that Agni will deploy his wealth in the arenas of clan and
house.
I.60.2: diváś cit pū́rvaḥ is standardly taken as temporal “before day(break),” but this
case form of div/dyu is more often spatial than temporal (note diváś cit … bṛhatáḥ in
the immediately preceding hymn, 59.5), and pū́rva- + ABL. has a spatial sense
elsewhere in just this ritual context: e.g., X.53.1 ní hí ṣátsad ántaraḥ pū́rvo asmát
“for he [=Agni] will sit down (as Hotar) close by, in front of us.” Cf. also II.3.3
mā́nuṣāt pūrvaḥ.
I.60.5: ab pátim ... rayīnā́m, prá śaṃsāmaḥ picks up 1c rayím iva praśastám, forming
a faint ring.
I.61 Indra
For general discussion of the intricate structure of this hymn, see the publ.
intro. as well as Jamison 2007: 60-68.
I.61.1: ṛ́cīṣama-, an epithet of Indra, is an impossible word; Ge wisely refuses to tr. it.
However, it is difficult not to see in it a combination of ṛ́c- ‘chant, song’ and samá‘like, same’, however obscure the morphological details are – and obscure they
certainly are. The first member cannot, straightforwardly, be a case form of ṛ́cbecause the case-ending should be accented. The length of the –ī- might be
analogical to the long ī in phonologically similar ṛjīpín-, ṛjīṣín-, but motivating a
short –i- (in putative *ṛ́ci-) is hard enough (Caland compounding form, like śúci-?).
It is tempting (and some have succumbed to the temptation) to connect –sama- with
another designation of ritual speech, sā́man-, but the difference in vowel length is
probably fatal. Note that in our passage the word is adjacent to another old crux,
ádhrigu-, the controversies about which (see KEWA and EWA s.v.) should have
been definitely settled by comparison with OAv. drigu- ‘poor, needy’ (Narten, YH
238–40). Both ṛ́cīṣama- and ádrigu- are disproportionately represented in the VIIIth
Maṇḍala, the home of much aberrant vocabulary. In the end those who elect to tr.
ṛ́cīṣama- take it as a compound of the two elements suggested above: Re ‘égal à la
strophe’, WG (somewhat peculiarly, though starting with the same elements) ‘der im
Preislied (immer) als dieselbe Person erscheint (?)’. For further, see EWA s.v.
I.61.1–2: Here and throughout the hymn, there is a certain amount of phonological
and lexical chaining (in addition to the repeated fronted demonstratives). Here 2b
bhárāmi picks up both 1c harmi and 1d bráhmāṇi, and 2a asmā́ íd u práyaḥ … prá
yaṃsi playfully echoes 1a asmā́ íd u prá…, with práyaḥ of 1b substituted for the bare
preverb prá.
77
I.61.2: In addition to the inter-verse echoes just noted, alliteration in 2c mánasā
manīṣā́ and 2d pratnā́ya pátye. JL adds 2a práya iva prá yaṃsi.
bā́dhe in 2b is universally taken as a dat. infinitive (as it is in I.132.5), but this
makes semantic difficulties because √bādh means ‘thrust, press, oppress’. Ge’s “um
(ihn) … zu nötigen(?),” Re’s “pour contraindre (le dieu),” and esp. WG’s “um (ihn)
… zu überhäufen” thus misrepresent the sense of the verb. The √bādh is esp.
common with the preverbs ápa and ví in the meanings ‘thrust away, thrust apart’, but
I here take the prá of pāda a with both verbs (bhárāmi and bā́dhe) in pāda b. As for
morphology, I take bā́dhe as a 1st sg. mid. pres. (the root is always inflected in the
middle), rather than as an infinitive, which allows the root to maintain its standard
sense: I push the hymn toward Indra with particular forcefulness. As a finite verb,
bā́dhe owes its accent to its initial position in the new clause.
suvṛktí- is a bahuvrīhi specialized for praise hymns (and occasionally the gods
who receive them) and is often simply tr. ‘praise(-song)’ (e.g., Ge ‘Preis’ here). I
prefer to render it literally; -vṛktí- belongs to the root √vṛj ‘twist’, and the English
idiom “good twist” refers to particularly clever turns in a plot or other verbal
products.
Most tr. take c with d, not ab. This is possible, but not necessary.
I.61.4: As Ge suggests (in n. 3 to his n. to 4ab), the apparent pleonastic doubling of
the simile particle (ráthaṃ ná tā́ṣṭā-iva) may instead signal that two images have
been crossed here: one with a simplex hinomi (“I impel the praise like a chariot”) and
the other with sáṃ hinomi and the addition of the carpenter as subject (“I, like a
carpenter, put together praise, like a chariot.”).
I.61.5: juhvā̀ has the standard double meaning, ‘tongue’ and ‘offering ladle’, a pun
that is enabled by the verb sám añje “I anoint”: anointing with the tongue means
producing praise, while ‘offering ladle’ fits better with the literal meaning of the verb.
dānaúkas- is likewise of double sense, both ‘accustomed to giving’ and
‘accustomed to gifts’, representing the reciprocal trade in praise and sacrifice given
to the gods, in return for the gods’ material gifts to us.
I.61.6: The tváṣṭā here has been prepared for by 4b táṣṭā, and both appear in
alliterative phrases: táṣṭeva tátsināya and tváṣṭā takṣat.
Another word with a standard double sense: ráṇa-; both senses are possible
here, also in vs. 9 below.
The position of yád in this subordinate clause is anomalous, as we expect at
most one constituent to precede the yá- form. I have no explanation, but there is
much that is off-kilter in the deployment of sentence parts in this hymn.
I’ve tr. the participle tuján as it were a finite verb, because the English
otherwise dribbles off into unintelligibility.
The unclear kiyedhā́- is found only here and in vs. 12. See EWA s.v. and
Scar’s (250–52) discussions of previous attempts at explanation. I favor the
78
suggestion registered (and dismissed) by Scar that it consists of the weak stem of
kíyant-‘how much, how great’ + the root noun dhā́-, with the development *-n̥t-dh- >
*-adzdh- > *-azdh- > -edh-, despite Hoffmann’s dismissal of the posited
phonological development (Aufs. 400), although I recognize the phonological
problems of this solution. Re’s “lui qui confère (on ne sait) combien” represents this
etymology one way or the other.
I.61.7: On the introductory gen. asyá referring to Indra, even though Indra is
otherwise in the nominative in this verse (as subj. of papivā́n [b] and of vídhyat [d]),
see disc. in intro. and in Jamison 2007 noted above.
The verse concerns Indra’s surreptitious drinking of his father’s soma right
after birth, an act enabled by his mother (ab), and Indra and Viṣṇu’s vanquishing of
the Emuṣa boar (cd), a rarely told tale. For Indra’s mother’s complicity in the somadrinking, see esp. III.48.2–3. In our passage Indra’s father is not directly referred to
(only by initial maháḥ ‘of the great [one]’ in b), but the beginning of the next word
promises the genitive of ‘father’ (i.e., pitúr), and only the final consonant of pitúm
removes that possibility — one of the many tricks Nodhas plays on us in this hymn.
(Ge ascribes the absence of pitúḥ to Worthaplologie, but I think rather that Nodhas is
laying a trap.)
On the basis of other tellings of the Emuṣa myth, the word to supply with
pacatám is odanám ‘rice-porridge’; cf. VIII.69.14, 77.6, 10.
The final pāda is quite artfully composed, beginning and ending with
alliterative phrases: vídhyad varāhám and ádrim ástā. Moreover, the first of these is a
variant of the very common formula that compresses the Vṛtra slaying, áhann áhim
“he/you slew the serpent.” Here, with the victim beginning with v-, the poet
substitutes a verb beginning with v-.
I.61.8: Although I tr. devápatnīḥ as ‘wives of the gods’, it is of course grammatically
a bahuvrīhi ‘having the gods as husbands’ (with the fem. stem patnī- substituted for
the masc. equivalent pati- in this feminine adjective). It reminds us of dāsápatnī‘having a Dāsa for a husband’, applied to the waters confined by Vṛtra, often
identified as a Dāsa, most famously in I.32.11 but also twice elsewhere of the waters
and Vṛtra (V.30.5, VIII.96.13). Since the context here is the Vṛtra battle (ahihátye),
the complementary terminology is probably deliberate.
JL notes that there is complementarity also in the second hemistich: 8c has
pári jabhre with Indra as subject and dyā́vāprthivī ́ as object, while 8d reverses this:
ná ... pári staḥ with dyā́vāprt̥hivī ́ as implicit subj. and Indra as implicit
object (via his mahimā́nam). This theme and its lexicon are picked up in the
9ab: mahitvám / divás pṛthivyā́ḥ páry antárikṣāt
I.61.9: The phrase “reverberant tankard” (svarír ámatraḥ) is striking, but intelligible
in Ṛgvedic context. Indra is compared to a large drinking vessel because of his
immense capacity and size, also described in 8cd and 9ab; “tankard” hints at his
ability to drink vast quantities of soma and implicitly promises his generosity
79
because he can contain vast quantities of goods. I follow Old in taking all forms of
ámatra- as belonging to a single stem (pace Gr, Lub, and EWA). As for ‘reverberant’
(svarí-), it echoes svarā́ḷ, which opens the preceding pāda; it also suggests the deep
sound made when a large (empty) vessel is struck and Indra’s own war-cries. (For a
possibly similar image, see I.100.12).
I.61.10: Numerous phonological plays in this verse: a śávasā śuṣántam; b ví vṛścad
vájreṇa vṛtrám; c gā́ ná vrā́nā avánīr [in this last example note the rhyme of the 1st
two words with the 2nd one].
I.61.11: This verse nicely juxtaposes a well-known deed of Indra’s, when he stops
the waters to make a ford for his client(s), with the even better known deed of
releasing the stopped waters in the Vṛtra myth, treated in the preceding vs. (10). On
the playful transition between these two myths here, see Jamison 2007: 113–14 n. 20.
The stopping of the waters causes mild surprise immediately after a verse concerning
their release.
tveṣásā is universally taken as belonging to Indra, whereas my published tr.
ascribes it to the rivers. I would now probably correct this, also to take it as Indra’s:
“just this one – with his turbulence/glittering,” because of the parallelism between
10a asyéd evá śávasā and 11a asyéd u tveṣáśā. But I am still disturbed by the form.
Its accent suggests that it should be adjectival, not a neuter -s-stem abstract like
śávas-, despite AiG II.2.224, which implies that it is nominal despite its accent. It is
curious that the expected root-accented neut. noun (*tvéṣas-) is not attested, and this
is the only (supposed) attestation of the suffix-accented stem. If it is a real adjective,
it could modify vájreṇa in pāda b, but this not only breaks the parallelism between
10a and 11a just noted, but, more seriously, would have to be extracted from one
clause and plunked into the next. (Ge in his note suggests supplying vájreṇa in pāda
a.) The a-stem adj. tveṣá- is also sometimes used of waters, e.g., VI.61.8 tveṣáḥ …
arṇaváḥ “glittering/turbulent flood,” which accounts for my original connection of
tveṣásā with the rivers.
The distribution of elements in pāda b is odd, with the subordinating yád in
normal 2nd position, but sīm, ordinarily another 2nd position element, just before the
verb.
īśāna-kṛ́t- is variously interpreted, either with the 1st member in a direct
object relationship with the 2nd -- “zum Herrscher, mächtig machend” (Scar’s tr; sim.
also Ge, Re, WG) – or in a sort of appositive subject relationship, “als Herrscher
handelnd” (Scar’s tr.; sim. Gr). Because the first member īśāna- is itself a participial
form ‘being lord, showing mastery’, I prefer the 2nd alternative. Note also that
independent ī́śana- is used three times of Indra in this hymn (6d, 12b, 15b), and it is
more likely that the same form in the compound refers to Indra’s masterful ways, not
to someone else whom he makes masterful. Scar allows both, though somewhat
preferring the 2nd.
More phonological play: c dāśúṣe daśasyan; d turvī́taye … turváṇiḥ.
80
I.61.12: On this very tricky verse, I simply reproduce (slightly paraphrased) my
discussion of it in Jamison 2007: 66: The beginning of 12 appears to return us from
the mythological past to the realm of the current-day poet of vss. 1-5; not only does it
start with the dative pronoun after a series of verses with the genitive phrase but it
continues with a standard lexeme for presenting a hymn to a divinity: prá √bhr̥:
asmā́ íd u prá bharā ... Cf. from the same poet also with a dative recipient I.64.1b.
nódhaḥ suvr̥ktím prá bharā marúdbhyaḥ “O Nodhas, present a well-turned (hymn) to
the Maruts.” But the phrase in 61.12 quickly goes awry. At the opening of the next
pāda, where Indra's name has been prominently placed in previous verses (índrāya
1d, 4d, 5b, 8b), we find, most shockingly, the name of his arch-enemy, vr̥trā́ya,
immediately followed by the accusative object vájram, Indra's weapon, not the word
for hymn we were expecting. The relevant parts of the half-verse asmā́ íd u prá
bharā..., vr̥trā́ya vájram... must mean "Towards just this one, towards Vr̥tra, bear
down the mace..." The poet has simply tricked us, having laid a trap with
conventional phraseology and syntax and with the stylistic patterns established
earlier in the hymn. He also skillfully exploits the morphological ambiguity of the
verb form bharā: given the pattern set in vss. 1-5 we are primed to interpret bharā as
a 1st sg. subjunctive (cf. indicative bhárāmi in vss 2-3), but as the half verse unfolds,
it becomes clear that bharā must rather be taken as a metrically lengthened 2nd sg.
imperative. [end of citation]
In order to let the audience in on the trick, the poet has imported much of
verse 6, the first mention of the Vṛtra conflict in this hymn: 6d reads tujánn ī́śānas
tujatā́ kiydhā́ḥ “gaining mastery, thrusting with the thrusting (mace), while
conferring (who knows) how much,” while 12ab echoes this with … tū́tujāno, …
ī́śānaḥ kiyedhā́ḥ “thrusting, gaining mastery, conferring (who knows) how much.”
In d íṣyan echoes ī́śānaḥ of pāda b (as well as 11c) and anticipates iṣṇānáḥ of
13c.
I.61.13: On this verse as a species of “poetic repair” of verse 13, see Jamison 2007:
66–67. The expected offering of praise to Indra thwarted in 12ab is successfully
effected in 13ab.
I.61.14: dyā́vā ca bhū́mā has the appearance of a dual dvandva, interrupted by the ca
that connects it to the NP with which it’s conjoined, giráyaś ca dṛḷhā́ḥ. But the
uninterrupted dvandva is actually dyā́vābhū́mī, with a different stem for ‘earth’, and
bhū́man- (n.) has no dual attested (and its dual should of course not be bhū́mā, but
probably *bhūmanī). See the next hymn for a variant on this usage. It is likely that
the lengthened N/A sg. form or the N/A pl. form is being used, but why? On this
problem, see AiG II.1.152.
In the publ. tr., “other” was careless omitted at the end of the first hemistich:
it should read “against each other.”
The mention of the poet Nodhas at the very end of the verse has perhaps been
prepared by several not entirely expected o’s: úpo ... jóguvāna oṇíṃ, sadyó, only the
last of which is a normal sandhi o < -as. The poet’s fondness for his own vowel may
81
account for the appearance of several relatively rare words: the intensive jógu- and
the noun oṇí-.
I.61.15: I follow Ge’s suggestion (in n. to 15a, followed by WG) that the unexpressed
subject that is being conceded to Indra is the soma-drink. The parallel he adduces,
V.29.5, contains the gods, the soma-drink, the concession, and Etaśa:
ádha krátvā maghavan túbhyaṃ devā́ ánu víśve adaduḥ somapéyam
yát sū́ryasya harítaḥ pátantīḥ puráḥ satī́r úparā étaśe káḥ
Then according to your will, o bounteous one, all the gods conceded the
soma-drinking to you,
when you put the flying golden mares of the Sun behind, though they were in
front, in Etaśa('s presence).
In the publ. tr. I take eṣām as a genitival agent, somewhat reluctantly. However, ET
suggests a much more attractive solution, which rescues the syntax: that eṣā́m is
dependent on asmaí “just to this one of them [=gods].” The singularity of Indra
would then be emphasized by the ékaḥ that begins the next pāda.
Etaśa is the horse of the Sun and is not, as far as I know, a soma-presser (the
physical image is a little comical). Against all tr., I therefore do not take súṣvim in d
as coreferential with étaśam in c, but rather as a second object with the verb. Nodhas
is now juggling the mythic past (the aiding of Etaśa) and the desired future (the
aiding of the soma-presser), as he nears the end of the hymn. I read the participle
paspṛdhānám twice, once with each object, with a different desired goal in the
locative for each. I also give the verb āvat two simultaneous morphological analyses,
fitting its two objects: the first as augmented imperfect to the Class I present ávati,
the second as perfect subjunctive to the pf. ā́va (cf. āpas, āpat taken by Hoffmann
1967: 64 n. 102, 101 n. 220, and Kümmel 2000: 118 as subjunctives to pf. ā́pa √āp
‘attain’). Although such an interpretation might be too artificial in some cases, I have
no hesitation in assuming Nodhas is capable of this.
Another phonetic play: d saúvaśviye súṣvim
I.61.16: hāriyojanā is unaccented and therefore taken by most as a vocative
addressed to Indra. I follow Old in emending to an accented form (hāriyojanā́),
modifying bráhmāṇi. See in the next hymn bráhma hariyójanāya “formulation for
the fallow-bay-yoking.” As Old points out, taking it as a vocative with Indra requires
either shortening the last vowel or allowing a lengthened vocative ending (which is
not usual), and possibly also shortening the first vowel.
This verse is extra-hymnic in some ways: its opening breaks the pattern set in
the previous 15 vss.; it makes a meta-announcement about the contents of the hymn
just recited; and it ends with the Nodhas refrain. But it also forms a ring with the first
verse, with the repetition of bráhmāṇi, and with the first part of the hymn concerning
the presentation of praise-hymns, with the reappearance of suvṛktí, which was a
signature word there (2b, 3d, 4c).
I.62 Indra
82
Shares much vocabulary with I.61 and sometimes plays off the turns of phrase
in that hymn.
I.62.1: Rich with phonetic figures: a śavasānā́ya śūṣám; b āṅgūṣáṃ gírvaṇase
aṅgirasvát [note that āṅgūṣám participates in both: it rhymes with preceding śūṣám,
but its initial matches aṅgirasvát, while gírvaṇase takes up the 2nd syllable of
aṅgirasvát]; cd ṛgmiyā́ya, árcāma arkám [this last also a clear etymological figure].
śavasānā́ya śūṣám recalls śávasā śuṣántam of 61.10a in the same metrical
position, with our phrase referring to Indra, the one in I.61 to Vṛtra.
The Pp. reads stuvaté in c, a reading that is universally followed. But this
dative is awkward: if it is coreferential with the other datives in the verse, it must
refer to Indra, and Indra “praising” is an odd concept in a hymn devoted instead to
presenting praises to him. Although Indra occasionally joins in the verbal parts of the
sacrifice (see above ad I.55.4), he does not ordinarily (or ever?) praise others.
Nonetheless, this is Re’s solution: “(dieu) louangeur.” Schmidt (1968, B+I, 163)
suggests a variant of this: “den mit guten Liedern (selbst) preisenden,” but this
suggests a medial form, not the active we have. If the participle does not modify
Indra, another person in the dative needs to be introduced, despite the unlikelihood of
a separate dative referent. This is Ge’s solution: “ihn [=Indra], der für den Sänger …
zu preisen ist,” construing stuvaté with ṛgmiyā́ya. So also WG with slightly different
tr., though both they and Ge consider the “praising” possibility. But the difficulty
disappears if, against the Pp., we read instead stuvatáḥ, genitive sg. of the participle,
as I do.
I.62.2: This verse is in some ways a double of vs. 1. They both begin with prá, and
the opening phrases prá manmahe and prá vo mahé rhyme, although the two mahe’s
have entirely different analyses. (Note also the mahé máhi play in 2a.) The next
phrase of 1a, śavasānā́ya śūṣám, is paralleled in 2b, with the first word identical and
sā́ma substituting for the second (both śūṣám and sā́ma referrring to the verbal
product offered to Indra). At the beginning of the second pādas, āṅgūṣyàm of 2b
matches āṅgūṣám of 1a. “Like the Aṅgirases” (aṅgirasvát) of 1b is picked up by the
Aṅgirases themselves áṅgirasaḥ in 2d, and the heavy etymology figure involving the
root √arc ‘chant’ is reprised by the participle árcantaḥ in 2d, which opens its pāda
just as the finite verb árcāma does in 1d.
I.62.3: We follow Janert (1956, Sinne und Bedeutung des Wortes “dhāsi” und seiner
Belegstellen im Rigveda und Awesta”) in taking dhāsí- as ‘wellspring’.
The post-caesura phrase in c bhinád ádriṃ vidád gā́ḥ contains rhyming verbs
followed by their objects; the disyllable ádrim contrasts with monosyllable gā́ḥ,
creating an almost syncopated effect. The strict parallelism of the two VPs may
account for the unnecessary accent on bhinát, which functions rhetorically like a
fronted verb, just like vidát. In fact, as my tr. indicates, I consider the initial subject
bṛ́haspátiḥ to be essentially extrasentential, a very topicalized topic.
83
The second of the VPs is also a mirror-image of the final phrase in 2d gā́
ávindan (imperfect in 2, versus aorist in 3), and it can be considered a “repair” of the
somewhat opaque vidát … dhāsím “found the wellspring” that intervenes (3b).
I.62.4: The repetitions and lexical and phonological echoes that have served Nodhas
well in the first three verses increase here, to an almost incantatory level. To point to
a few of these: a sá suṣṭúbhā sá stubhā́ (again a syncopation, as in 3c); b svaréṇa …
svaríyo …; c: saraṇyubhiḥ echoes the sá’s of a and the s(v)ar’s of b, while finding its
own faint echo in the final śakra of the pāda; d ráveṇa is a virtual anagram of
svaréṇa in b, darayo is almost a rhyme form of svaríyo in b in the same metrical
position, and both svaríyo and darayo precede the hemistich-final words ending in gvaiḥ.
All other tr. take this verse as a single sentence with 2nd ps. subj. However, as
I have shown (1992: sa figé), sá with 2nd ps. reference is limited to imperatives and
should not appear with an injunctive like darayaḥ. I therefore consider the first
hemistich to continue the 3rd ps. reference of vs. 3, with a switch to 2nd ps. in the
second half. This type of switch is quite common in the RV.
In c I do not take saraṇyú- and phaligá- as personal names, pace Ge et al. The
tr. ‘bolt’ for the latter follows Hiersche (Asp., ‘Riegel, Vershluss’), based on a
possible connection with parigha- (Up+) ‘iron bar for shutting a door/gate’; see
EWA s.v. However, as ET points out, a less specific sense such as ‘barrier’ would fit
the contexts better and would also make fewer assumptions about early technology.
I.62.5: The instrumentals in b (uṣásā sū́ryeṇa góbhiḥ) are not parallel with
áṅgirobhiḥ in a. The Aṅgirases there are the agents of the passively used participle
gṛṇānáḥ (and also potentially instrumental of accompaniment with the main verb, as
Indra’s helpers: “along with the A’s you uncovered…”). The instrumentals in b are
the additional elements that Indra uncovered.
ándhaḥ is a potential pun, as a homophone meaning both ‘blind darkness’ and
‘soma stalk’. The former is surely the first reading in this treatment of the Vala myth,
with the cave a black hole, as it were. So Ge, Re, WG. But soma is never far from
Indra’s mind, and the cosmogonic deeds with which he is credited in the second
hemistich are often performed under the influence of soma. Schmidt (p. 164) favors
‘(Soma-)Flut’, following Bergaigne and Lüders.
I.62.6: This verse has to be Nodhas’s joke. It is conspicuously placed, as the central
verse of the hymn and thus a potential omphalos, and it announces Indra’s deed
(kárma) with extraordinary fanfare, including two superlatives (práyakṣatamam …
cā́rutamam). But in a hymn so far devoted to what is one of Indra’s greatest deeds,
the opening of the Vala cave, the poet springs on us instead a deed of utter obscurity
involving the swelling of four rivers, an act with no other clear mentions in the RV.
Ge valiantly seeks parallels, but the two passages he adduces (I.104.3–4 and I.174.7)
have little or nothing in common with our passage and it’s not even clear that there
are rivers in the second. Lüders (Varuṇa 335-37) predictably sees these as heavenly
84
rivers – four because they flow “nach den vier Himmelsgegenden.” Re also sees
them as associated with heaven, while Schmidt B+I 164) and WG seem to see them
as being “under” (something unspecified). I have no solution for the affinities of this
scrap of mythology, although I tend to agree generally with Lüders that the rivers are
more likely to be celestial than terrestrial, and I would suggest that it may have to do
with producing rain. But I still think the point here is that Nodhas has set us up for a
grand announcement and then, by bait and switch, given us a myth that none of us
has ever heard of.
The asti in this hemistich may provide another bit of evidence for this view.
Unaccented asti almost always has existential value (“there exists…”) because the 3rd
sg. pres. copula in equational sentences (“X is Y”) is regularly (indeed, probably by
rule) gapped. But an existential sense here doesn’t work (“There exists this most
conspicuous deed…”). I suggest that the overt asti here signals a strong assertion in
the face of expected opposition (“This is his most conspicuous deed” – rather like
American children’s quarrels: “is too” “is not”).
The disputed word upahvaré adds to the obscurity of this bit of myth. It is
clearly a derivative of √hvṛ ‘go crookedly, deviate’, but the exact sense of this
nominal derivative is not clear. In this passage Ge takes it as ‘lap’ (Schoss), Re as
‘fold, crease’ (repli), WG as ‘abyss, chasm’ (Abgrund), while Schmidt bypasses any
literal rendering with “im Felsen.” Lüders considers it the name “für den Behälter
des himmlischen Urquells,” which of course fits his larger picture of the heavenly
ocean. I consider the word to convey in the first instance a visual image, that of a
meander or deviation from the straight. In connection with rivers (as also in
VIII.96.14) it refers to eddies, the circular or oval shapes produced against the
current by uneven flow. In other geographical contexts it can refer to byways,
detours from the straight path, and ultimately to remote places, the backeddies, as it
were, of the mountains.
I.62.7: Nodhas continues to deploy his tricks in this verse. He has moved on from the
obscure four-river reference in the preceding verse, to a more standard domain for
Indra to display his power: the two world halves. But in the first hemistich Indra is
not depicted as filling them or propping them apart or any of his usual actions with
regard to them, but rather as “uncovering” them, using the same lexeme ví √vṛ as
was just used in a Vala-myth verse (5a ví var, 7a ví vavre). (That lexeme is close to
being the signature Vala verb, though it is actually more commonly ápa √vṛ, with the
same sense, but a different preverb.) Notice that the two occurrences of ví √vṛ
surround the pseudo-omphalos verse 6, in the usual manner in which concentric
lexcial rings signal an omphalos. So Nodhas thus presents this new mythological
theme as if it were the old one, the Vala myth found in vss. 2-5, though he is entering
different mythological territory.
He also reuses the √ṛc lexical theme from vss. 1–2, with arkaíḥ in b, which is
apparently qualified by the middle participle stávamānebhiḥ (remember stuvatá in
1c). This participle is somewhat disturbing; like all middle participles to the root
√stu, it is used passively here, but unlike all the other passive uses, its subject is not
85
the god praised but the praises themselves. It may be that Nodhas want his audience
to take notice of such an aberrant usage. However, I am tempted to read
*stavamānaibhíḥ here – that is, nom. sg. stávamānas + pron. ebhíḥ, with rare double
application of sandhi: -as +e- > -a + e- > -ai-. The participle would modify Indra, as
would be expected; ebhíḥ occurs several times with arkaíḥ (IV.3.15, 10.13) and the
passive of √stu also appears with this instr. (see passages adduced by Ge in n. 7b,
where he suggests a similar, but far from identical, reanalysis of the form in the text,
calling it “vielleicht Hypallage.” If the emendation is accepted, I would change the tr.
to “the irrepressible one being praised by these chants.” The arkaíḥ might also be
construed secondarily as it currently is in the tr., with ví vavre, as the instrument with
which he performed the uncovering.
ayā́sya- regularly qualifies Indra; there is no reason to take it as a personal
name, identifying a distinct second figure here (pace Ge, Re).
Phonetic figure: sanájā sánīḷe taking up the sa’s in vs. 4 and prefiguring the
sa’s that will be conspicuous in the next few verses.
The second hemistich contains a clever pun between simile and frame. In the
frame (pāda d) Indra supported, that is, held up, the two world halves in familiar
fashion (somewhat repairing the less standard uncovering he performed in ab, though
cf. VIII.96.16). In this task he is compared to the god Bhaga (though it is not an
activity that I think of as particularly associated with Bhaga). But in the simile
bhága- is used as a common noun ‘(good) portion’, which provides support for two
consorts (méne) – the point presumably being that a man needs a particularly large
portion of property to support two wives. I do not follow Hoffmann’s view (1960:
245; KZ 76) that méne here is an elliptical dual referring to concubine and lawful
wife, and in fact believe that ménā in general has a wider sense than he allows. He
restricts it to concubines, but the derivation from √mi ‘exchange’ suggests a wider
application, with the wife as an “exchange token” in the economic transaction of
marriage.
I.62.8: From the spatial dual females of vs. 7 Nodhas now moves to the temporal:
night and dawn, whose activities have kept occurring “from of old” (sanā́t), linking
them lexically to the two world-halves of 7a who were born of old (sanájā).
In dívam pári bhū́mā we encounter the not-quite-dvandva also met in 61.14;
there in dyā́vā ca bhū́mā, dyā́vā was definitely dual, but its partner bhū́mā was either
singular (with lengthened final) or plural. Here dívam is definitely singular, with
bhū́mā either of the choices just given, probably singular.
punarbhū́- later (already in the AV) has a specialized legal application as a
designation for a remarried widow (see Thieme Kl.Sch. 445-60 = KZ 78, 1963), but
this sense does not seem to be in play here or in the other RV passages.
Though most take aktā́ (which only appears here) as the ppl. of √añj ‘anoint’,
serving as a poetic designation for ‘night’, I favor a root connection with the ‘night’
word, with zero-grade in the root syllable, esp. since aktú- ‘night’ is probably so
connected (pace EWA s.v.). See EWA s.v. and AiG III.234, though both are
skeptical. Of course secondary association with √añj would be natural.
86
JL notes the chiasmic kṛṣṇébhir aktā́-uṣā́ rúśadbhiḥ of c; for a similar
chiasmus see 11c.
JTK suggests that vápus- here and in other dawn/night passages refers to
bridal clothing and suggests a tr. ‘finery’.
I.62.9: Although the first hemistich is perfectly easy to tr., it is not at all clear what it
is referring to. First, what is the “partnership” (sakhyám)? Ge thinks it is with the
beings named in vss. 7-11, while Sāy thinks rather of the sacrificers. Nothing in the
context helps to define what type of partnership it is or with whom, and the verb that
governs it, “supported” (dādhāra), seems oddly chosen. The subject of this verb is
also not fully identified: the larger context of course points to Indra, and sudáṃsāḥ
‘possessing very wondrous power’ was just used in 7c of Indra (where he was
likewise the subject of a form of √dhṛ), and see dáṃsaḥ applied to his deed in 6b.
But who is he the “son” (sūnú-) of? and more to the point, the epithet śávasaḥ sūnúḥ
“son of strength” is ordinarily Agni’s. Here we have the variant sūnúḥ … śávasā
with instr., and sūnúḥ śávasā is used of Agni in I.27.2. On the other hand, śavasānáis twice use of Indra in the beginning of this hymn (1a, 2b) and śavasāvan in 11d,
śavasāna in 13c; cf. also the clear śávasaḥ sūnúm índram in IV.24.1. I tentatively
suggest that there is an attempt to blur the lines between Indra and Agni here, and
that the partnership or fellowship may be the mutually beneficial relationship
between gods and men centered on the sacrifice, where Agni officiates. But I cannot
point to hard evidence for this interpretation.
The paradox of the cooked milk in the raw cows is a standard one, often
trotted out as a natural wonder to contemplate. Here it seems to be taken further:
there is literal milk and literal cows in c, but in d the black and red cows are likely to
be the nights and dawns (as in 8cd) and the “gleaming white milk” the early light
known as the “milk of the dawn cows” (see Watkins MoreDawnCows). As far as I
can see, this hemistich has no integral connection with the first half of the verse, and
the shift from 3rd to 2nd person underlines this disconnect.
I.62.10: A nicely balanced alliterative pāda sanā́t sánīḷā avánīr avātā́, with the first
phrase picking up 7a sanájā sánīḷe. Note also that the first word of b, vratā́, echoes
avātā́. On avātá- see comm. ad I.38.7 and VIII.79.7.
Again the reference of this verse is somewhat obscure. Are the streams here
the same as the four rivers in 6cd? or are they the streams released from Vṛtra’s
confinement? (Ge suggests sensibly that Indra gave them the command to wander in
I.61.12, though it is their own commandments, not Indra’s that they safeguard here.)
The identity of the sisters of cd, compared to wives, is even less clear. Are they the
same streams as ab? or possibly praise-songs (gíraḥ appear with the same simile,
jánayo ná pátnīḥ in I.186.7, and in our next vs., 11cd, “inspired thoughts” [manīṣā́ḥ]
are compared to eager wives). Again, I think the reference is supposed to be shifting,
with the literal streams of mythological time replaced by the thoughts and praises of
the ritual here-and-now, mediated by the feminine gender of the words used of both
streams and praises. The temporal shift in this part of the hymn, or rather the double
87
temporal vision, is signalled by the repeated initial sanā́t ‘from of old’ and related
terms (8a, 10a, 11a, 12a, 13a) and its contrast with the present tense verbs that
prevail in this section.
I.62.11: sanāyú- ‘age-old’ is found only here, and was clearly created to match the
other pāda-initial sanā- ‘old’ forms in this section, while matching its nom. pl.
sanāyúvo with vasūyávo, which opens the next pāda. JL suggests that part of the
motivation for creating this particular hapax is to invoke the paradoxical
juxtaposition of sana- ‘old’ and yuvan- ‘youth’. Verse 8 contains the same
juxtaposition: sanā́t … yuvatī́ “from of old the two young girls,” and the theme of old
and new runs throughout this hymn, indeed in this verse.
JL also notes the chaismic pátiṃ ná pátnīr uśatī́r uśántam of c.
I.62.13: sanāyánt- is likewise only here. Note its semantic opposite návyam at the
end of the pāda, the same contrast as sanāyúvo … návyo in 11a.
Like the last verse of the preceding hymn (I.61.16), this verse is extra-hymnic
in some ways, but also has connections with the rest of the hymn. The “old” pattern
continues to open the verse, and the voc. śavasāna in c forms a ring with the same
stem in the dative in 1a.
I.63 Indra
I.63.1: The very first hemistich of this hymn presents us with a syntactic problem:
although the most natural way to read this is with a rel. clause beginning with yó ha
and continuing through the second pāda, the verb there, dhāḥ, is unaccented. There
are three possible ways to handle this. The first, the one I followed in the publ. tr., is
to assume that dhāḥ got de-accented for some reason. Several possible reasons are
available, though none of them strong: 1) it lost its accent redactionally from being
matched to (properly) unaccented dhāt in 2b, also at the end of its hemistich; 2) the
same idiom áme (…) dhāt is found in nearby I.67.3 (unfortunately by a different
poet) and this caused the loss of accent redactionally; 3) because the 2nd hemistich of
the verse in question is also a subordinate clause (beginning yád dha), the previous
subordinate clause lost its verbal accent and became a kind of “honorary” main
clause, because the action of cd is based on that of the relative clause that precedes it.
(This seems to be one possibility that Old floats, though he is rather sketchy about it.)
Though Ge does not mention the lack of accent, he tr. the whole as a rel clause.
The other two ways to deal with the problem involve having the rel. clause
end at the end of the first pāda. Old also suggested the possibility that the rel. cl. is
limited to yó ha śúṣmaiḥ. Re, Hoffmann (1967: 181), and WG all follow this path,
with tr. of the type “You are great, Indra, who are (so) because of your śúṣma.” I
rejected that because of its awkwardness, although I recognized that it solves the
accentless verb problem, since the main clause can occupy pāda b without problem.
A third way, also ending the relative clause with the end of the first pāda, seems to
me to be less clumsy, and that is to make all of the first pāda the relative clause.
88
Though yáḥ appears to be fairly far into that clause, in fact it is not preceded by more
than one constituent (plus a vocative, which doesn’t count). This produces a tr.
“(you) who are great because of your śúṣma-, you placed ….” This actually has a
parallel in IV.22.3 yó deváḥ …, mahó vā́jebhir mahádbhiś ca śúṣmaiḥ “The god who
… is great by reason of his prizes and his great śúṣma.” I would therefore now
change the tr. accordingly, to this third option.
The next problem in this half verse is the construction of áme √dhā. All the
standard tr. render it more or less as “you put Heaven and Earth in panic/fear.” This
makes sense of the syntax, but attenuates the meaning of the noun: áma- is an attack
or onslaught, or more abstractly offensive power, whereas ‘panic, fear’ suggests the
opposite. Therefore, although slightly more machinery is required, I tr. this idiom
(found also in I.67.3, IV.17.7) as “put X in the path/way of your onslaught.”
Notice that here Nodhas manages to deploy a perfectly orthodox de-coupled
dual dvandva for Heaven and Earth (dyā́vā … pṛthivī́). See remarks on I.61.14, 62.8.
The subject NP(s) of cd are neatly interwoven: with víśvā … ábhvā (neut.)
chained with giráyaś cid … dṛḷḥā́saḥ (masc.).
Hoffmann (p. 181) argues convincingly that naíjan should not be interpreted
as ná aíjan with the Pp., but with the injunctive éjan.
I.63.2: If my new parsing of the syntax of vs. 1 is correct, these two verses are
structurally parallel, with a relative/subordinate clause in a, the main clause in b, and
a further subordinate clause, esp. dependent on the main clause, in cd.
Note the phonological trick, where (ā́) yád dhá(rī) opens 2a, apparently
matching yád dha of 1c (which in turn shows dha for ha, found in the grammatically
different rel. clause yó ha in 1a). The verse ends with more alliteration: d púra(ḥ)...
puru- ... pūrvī́ḥ.
The meaning of vívrata-, several times of Indra’s horses, but with a few other
referents, isn’t entirely clear, because it depends on which of the many value of ví to
assume. I favor ‘having separate commandments’, that is, acting independently but
each in obedience to Indra, but it could also mean ‘without commandments’ (which
must underlie Ge’s ‘widerspenstige’ (stubborn), followed by Hoffmann (but with ?).
The c pāda lacks a caesura and in its place has a somewhat bizarre compound,
aviharyata-kratu-. The second member is clear, and the whole must be a bahuvrīhi.
The first member appears to contain the well-attested adj. haryatá- ‘gladdening’, but
this presents certain problems: the present háryati does not appear with the preverb
ví and it is difficult to construct an appropriate meaning, esp. one that would
harmonize with krátu- ‘resolve’. Possibly, with double negation via a + ví, “whose
resolve is not undelightful” – not a particularly compelling possibility. Nonetheless,
this derivation seems the path of least resistance and is probably responsible for Ge’s
“dessen Rat nicht verschmäht wird” and WG’s “… nicht unbeliebt ist” (possibly also
Hoffmann’s “… nicht zu unterschätzen ist” – it’s hard to tell). But all of these must
take the -haryata- in a passive sense, although the simplex adjective never has that
value.
89
I think it’s likely that the compound involves some complex word play,
scrambling and recombining the (d)hárī and ví(vratā) of pāda a. Despite the ad hoc
explanations required, I’m inclined to consider the –viharyata- somehow a derivative
of ví √hvṛ ‘deflect, swerve, overturn’, a well-established idiom. (Re’s “dont le
pouvoir est incontrariable” seems to reflect this.) Two negated nominal derivatives of
this lexeme are found in the RV: ávihruta- ‘unswerving’, jpb: ‘not overturned’
(V.66.2, X.170.1) and ávihvarant- jpb: ‘not overturning’ (IV.36.2). Unfortunately the
morphological details are not going to be easy. Accounting for the loss of the v in a
putative *avihvaryata- is not so difficult: it can be ascribed to dissimilation, in
combination with the word play on (d)hárī just mentioned. But what about the rest of
the word. For the suffix –ata- I can only invoke the (small) class of –ata- adjectives
with full-grade root syllable of the type darśatá- ‘(lovely) to be seen’, yajatá‘worthy of the sacrifice’, and the aforementioned haryatá- ‘delightful’. As for the
intermediate –y- between root syllable and ending, I again invoke the play on (d)hárī,
with the –y- representing the sandhi of –ī- before the suffix –ata-. If word play is
indeed involved here, it’s possible that a second reading of the first member could be
“not without his fallow bays” and the whole compound means “who has no resolve
without his fallow bays.”
I am fully aware of what a house of cards this is and only sketch these
scenarios because I think Nodhas is capable of such manipulations. I do think it
possible, however, that the lack of caesura and the problematic compound are
connected and that the transmitted text is not what Nodhas originally composed. It
should be noted that the word division yénā viharyatakrato is also a possibility, but
this doesn’t help either with the lack of caesura or the analysis of the compound.
I.63.3: I interpret dhṛṣṇúḥ as governing the acc. etā́n, though as a goal not a direct
object..
I.63.4–7: tváṃ ha tyád provides the opening of the next four verses. Although tyád is
a neut. pronoun and in 4a could function as the object of the otherwise object-less
codīḥ, in the following verses it does not have pronominal function; the expression X
ha tyád is a way of providing emphasis, particular of the 2nd ps. pronoun (cf., e.g.,
VII.19.2, VIII.96.16–18).
I.63.5: This verse contains many small problems of interpretation that hinder putting
together the whole. I’ll start with the details and turn to the construction of the whole
thereafter.
áriṣaṇyan is universally taken as intrans., patient-oriented ‘not receiving harm,
not being harmed’, but in all cases this stem is better taken as ‘not allowing harm’.
The finite verb riṣaṇyáti also has this meaning “intends harm, allows harm’. The
parallel stem áriṣaṇya- in II.29.4 has the same meaning and takes a genitive
(áriṣaṇyā tanū́nām “not allowing harm to our bodies”), and I construe the gen.
mártānām in b with áriṣyan: “not intending harm to mortals.”
90
The next problem is dṛḷhásya. Ge takes this as the causer of the harm to Indra
(gen. agent), while Hoffmann, followed by WG, construe it with ájuṣṭau, with the
gen. pl. mártānām dependent on dṛḷhásya (KH “bei Missfallen selbst eines
Feststehenden unter den Sterblichen”), but both the Ge and the WG interpretations
require that dṛḷhá- be a person. This is unprecedented: this past participle is virtually
always used of a place, a fortified stronghold. Indra is once (VIII.44.12) called dṛḷhá-,
presumably because he is as steadfast as a fortress (in the vein of “a mighty fortress
is our god”), but I know of no examples with people so called. If at all possible it
should have the standard meaning here. I think this is possible, by reading the verb ví
… var of pāda c also with the first hemistich and supplying the word dúraḥ ‘doors’.
In fact, the other two passages containing the gen. sg. dṛḷhásya strongly support this
solution, as they both contain the verb ví √vṛ and the headnoun dúraḥ: VI.62.11
dr̥ḷhásya cid gómato ví vrajásya, dúro vartam... “Open the doors of the cattle pen,
even though it is shut fast” and VII.79.4 ví dr̥ḷhásya dúro ádrer aurṇoḥ “you opened
up the doors of the firm-fixed stone.” Note that ví … var is rather more appropriate
to this hemistich than it is to the pāda in which it’s found. (Recall the double use of ví
√var in I.62.5 and 7.)
The last word to be accounted for in this hemistich is ájuṣṭau. Unlike most tr.
I take it to be Indra’s displeasure. The object of his displeasure is not expressed, but
it is easy to supply – presumably Vala, since Indra is opening up his stronghold. His
displeasure at Vala (or other demonic enemies) is implicitly contrasted with his
benevolent care not to injure mortals.
I am less sure about what pāda c is conveying. I assume this is a racing image:
Indra starts the race by opening the barriers at the starting line. Ge and Re take the
injunctive ví … var as a functional imperative, but this is certainly not necessary. I tr.
as a past tense, in keeping with its rendering in the first hemistich, but it would be
possible to take it as a present in c, as Hoffman/WG do. The phrase asmád ā́
“(coming) from us” seems a bit heavy simply to express that it’s our horse, so there
may be some racing terminology embedded here.
In d ghanéva contains the old instr. ghanā́.
I.63.6: The gerundive atasā́yya- is taken by Ge as belonging to √at ‘wander’ and tr.
“soll … erreichbar sein,” while most (Gr, Re (?), AiG II.2.286, Hoffman, WG) take it
to mean “to be asked for/begged” and derived from the hapax atasī́- supposedly
‘begger’ (VIII.3.13). But this is entirely circular, with the meaning of each of these
two words determined on the basis of the other, with an occasional nod towards √at
‘wander’ as their possible root (whose relevance Mayrhofer casts doubts on; see
EWA s.v. atasī́-). I instead take it as a negated gerundive to √taṃs ‘tug, yank, shake’,
which forms a similar gerundive to its intensive, vitantasā́yya-. Negated gerundives
do not ordinarily take accent on the privative (AiG II.1.217), and the derivational
opacity of atasā́yya- would make it likely to receive the usual –ā́yya- gerundive
accent. Ge also suggests the possibility of derivation from √taṃs in a note.
91
I.63.7: “one after the other” is my attempt to render an “object-distributive” value for
the intensive dardaḥ (for this notion see Schaefer 1994: 86–87, though she considers
dardar- lexicalized and does not attribute this sense to it, pp. 135–36).
The use of várg ‘you twisted’ differs in tone in similar and frame. In the
frame it is distinctly hostile and destructive, but in the simile it simply expresses the
ritual action of twisting the barhis grass into seats for the visiting gods.
I.63.8: ā́paḥ in b can certainly be the old nominative plural used as an accusative, as
occasionally happens in late RV (so Gr and implicitly most tr.), but as I suggested
elsewhere (1982: 260) it may be the nominative plural it appears to be; this may be
an example of case disharmony in a simile, exploiting the ambiguities in syntax of
the verb √pi ‘swell’ that the poets were always happy to capitalize on.
I take tmánam as functioning as a reflexive pronoun (“extend yourself”),
while most tr. take it as a noun (Ge ‘Lebensgeist’, WG ‘Lebenshauch’). This is the
only acc. to the truncated stem tmán-, and so its usage is hard to determine. The
oblique cases (dominated by inst. tmánā) generally refer to the self (‘by yourself, for
ourselves’, etc.), without a full-blown nominal sense, which accounts for my tr.
However, the active form of yáṃsi gives my pause, and so the standard tr. should be
considered (though I am not ready to adopt it). “Extend nourishment to us” seems to
be formulaic, with an active form of √yam, and so that voice may have prevailed
here, despite the reflexive tmánam; cf. III.1.22 prá yaṃsi … íṣo nah, IV.32.7 sá no
yandhi … íṣam.
I.63.9: I supply ‘hymn’ as subject for ákāri in the first pāda and take bráhmāṇy ùktā
in b as a separate nominal clause. It is of course grammatically possible to take the
latter phrase as subject of ákāri, since neuter plurals famously can take singular verbs.
But, fame aside, this construction is quite rare in the RV, and in this case there are
two different recipients, Indra in a and the horses in b. However, I do not consider
the other interpretation impossible.
I.64 Maruts
I.64.1: Ge argues that apó (apparently the acc. pl. of ‘waters') should be emended to
ápo (neut. acc. sg. of ‘work’). His parallels are persuasive, and I accept the
emendation. Therefore the published tr. should have an asterisk * before ‘work’. The
accent shift may have been made redactionally, on the basis of vs. 6, which contains
apó ‘waters’ in its first pāda and vidátheṣv ābhúvaḥ in its second, as here. On the
other hand, if we choose to retain the reading apó (see below), the hemistich can
mean “Wise in mind, with dextrous hands I anoint the hymns, like waters standing
ready at the rites.” WG accept the reading as given, but take it as “Werkmeister,” i.e.,
to the internally derived adjectival stem apás- ‘possessing work’. This would be
appealing save for the fact that the masc. nom. sg. should be *apā́s (or ápā in this
sandhi context). This morphological problem is not mentioned.
92
I.64.4: vy àñjate here responds to sám añje in 1d, though here the verb is reflexive
and in the 1st vs. it is transitive but with subject involvement (“I ornament [my own]
hymns”). I tr. the verbs slightly differently because the ‘anoint’ sense doesn’t work
very well in vs. 1, esp. with the simile, whereas here the etymological figure with
añjíbhiḥ ‘unguents’ enforces the more literal sense.
Gr suggests emending mimṛkṣur ‘they rubbed’ to mimikṣur ‘they were
fixed/attached’ (accepted by WG), since this verb √myakṣ is used elsewhere of
spears in similar contexts. However, ‘rub’ makes fine sense, and the poet may well
be playing with a well-known phrase. (This perhaps should give us pause about
emending apó to ápo in 1.)
I.64.5: Very alliterative verse, esp. in 2nd half, with pāda c duhánty ū́dhar divyā́ni
dhū́tayo (cf. also dhúnayo in pāda a) and d … pinvanti páyasā párijrayaḥ. Cf. also b
vā́tān vidyútas (ta)vi(ṣībhir).
Most tr. take ū́dhar divyā́ni together (e.g., Ge “die himmlischen Euter”).
Although divyá- does modify ū́dhar- elsewhere in similar context (e.g., IX.107.5
duhāná ū́dhar divyám mádhu priyám) the collocation is not grammatically possible
here, since ū́dhar is definitely singular and divyā́ni is definitely plural. At best one
would have to argue that short-vowel neut. plurals to n-stems, which are identical to
their neut. sgs., provided some sort of model (cf., e.g., I.173.3 sádma mitā́ “fixed
seats”). Old (ZDMG 55.273 n. 1 [=KlSch 732]) claims that this is a vicarious
introduction (“ein vicarirendes Eintretung,” whatever that means) of the sg. ū́dhar
since the plural of ū́dhar- doesn't exist. (We would expect *ū́dhāni, like áhāni to áhar,
though Old bizarrely produces rather *ū́dhṝṇi.) But the issue here is not the singular
of ‘udder’, but the plural of ‘heavenly’, which the poet could easily have made
singular (as in the parallels) if he had wanted. The root √duh can take a double
accusative, as in IX.107.5 just quoted: “milking the heavenly udder of its dear
honey”), and this is how I interpret divyā́ni here, supplying a form of vásu ‘goods’,
which divyá- regularly modifies. In one way or another this alternative goes back to
Gr and Delbrück AiS.80 and is accepted by WG. Again, Nodhas seems to be
playfully evoking a familiar phrase but modifying it.
I.64.6: vidátheṣv ābhúvaḥ, repeated, as noted above, from 1d, is ambiguous. It can
modify the waters of the first pāda, as Ge (/WG) takes it, or the Maruts (so Re). I
have tried to maintain the ambiguity. In vs. 1 the phrase qualified the hymns (fem. pl.
gíraḥ) and, if apó of pāda c is not emended, those waters as well.
I.64.8: piśá- in b occurs only here in the RV, so although the meaning assumed by
most (‘mottled [deer]’ vel sim., beginning with Sāy) is plausible, it has no other
support. The plethora of wild animals in this run of verses (7: buffalos, wild
elephants, 8: lions, snakes) does suggest that this term names animals too. Re by
contrast tr. the etymological phrase piśā́ iva supíśaḥ as “bien ouvragés comme des
oeuvres-peintes,” though he is more cautious in his n. Kulikov (JIES: 2009) suggests
‘cheetah’ or ‘leopard’ on the basis of the wild animal company it keeps in its AV
93
occurrence (AVŚ XIX.49.4 = AVP XIV.8.4). This is plausible but by no means
certain.
I.64.9: As noted in the intro., ródasī is a pun. In pāda a it refers to the two worldhalves, but in cd the unexpressed subject is feminine and mounts the Maruts’ chariot.
This can only be Rodasī, the Maruts’ consort (for mounting their chariot, see
I.167.5–6), who ordinarily is accented rodasī́ but here must be generated by
otherwise homonymous ródasī in pāda a.
I.64.11: payovṛ́dh- has several possible values. In addition to the ‘milk-strong’ in the
publ. tr., it would also be possible to take the -vṛdh- transitively: “strengthening with
their milk [=rain]” or “strengthening the milk [=rain]”; either alternative is supported
by vss. 5–6.
The lexeme úd √han is quite rare and occurs with the verb stem jíghna- only
here. Since the exact sense of the lexeme is not certain, it is difficult to know how to
complete the simile. I have interpreted the verb to mean ‘beat up’; with regard to
roads there is a nice parallel in the English idiom “pound the pavement.” I assume
that ā́pathi- refers to walkers on a path and supply ‘road’ as the object parallel to the
mountains and their feet as the instrument of pounding that is parallel to the wheelrims. ET compares the French idiom ‘frayer le chemin’, used of blazing a trail,
which may well be the sense meant here. Ge takes the verb as ‘aufreissen’ (tear
open/up) and supplies ‘stones’ as the object (sim. WG). I am a little puzzled as to
why his Strassenfahrer would be tearing up stones, but I suppose it might be to clear
a rugged path. Re: “… arrachent les montagnes, tels des frayeurs-de-voie.” I do not
know what he means by this.
I.64.12: The word sūnú- ‘son’ in the singular takes one aback in a hymn devoted to
celebrating the very plural Maruts. It clearly refers to the ‘flock’ (gaṇá-) in pāda c,
but I have tr. ‘offspring’ to avoid the misleading “we hymn the son of Rudra…”
I.64.13: Ge (/WG) construes āpṛ́chyaṃ krátum with ā́ kṣeti, giving the latter the
meaning ‘possess’. But this Class II present means only ‘inhabit, dwell’ (as opposed
to Class I kṣáyati ‘rules over, owns’ to a different root √kṣi). Re clearly recognizes
the problem and tries to reconcile the meaning of the root present with its construal
with an object, producing the very odd “il habite une force-spirituelle digne-d'etrerecherchée.” It seems preferable to me to take āpṛ́chyaṃ krátum with the previous
pāda and tr. ā́ kṣeti púṣyati separately. The two verbs (admittedly without the preverb
ā́) form a fixed expression (I.83.3, V.37.4, VII.32.9). The position of the preverb also
does not favor construing the preceding accusative with ā́ kṣeti: we might then expect
the preverb to begin the clause, though this argument is considerably weakened by
the fact that the pāda would begin *ā́ āpṛ́chyam, in sandhi ā́pṛ́chyam, and the separate
preverb would be essentially lost. As for the sense of bharate … āpṛ́chyaṃ krátum, I
interpret it as what I think Re would call a “legère zeugma.” The first two objects of
bharate, vā́jam and dhánā, are both physical objects that the subject carries from one
94
place to another; the third proposed object is a mental attitude. The English idiom
“carries himself” and the English noun “bearing” use ‘carry, bear’ in a similar sense,
to refer to a person’s mental self-presentation, and the middle voice of bharate would
neatly fit such a value.
I.65 Agni
I.65.2–3: These verses contain 3rd pl. act. root aorists ánu gman (2b) and ánu … guḥ
(3b) to the associated roots √gam and √gā respectively, with no apparent difference
in semantics. Here one might almost invoke the much over-used explanation of
metrical convenience: both of these 3rd plurals are always (gman) or almost always
(guḥ) pāda-final, with the first “making position” with a preceding light syllable
(often ánu) and the latter not.
I.65.3: The second half of this verse is very economically and ingeniously expressed.
First, the syntactic modes of simile and frame are entirely different. The frame
consists of a single noun, nominalized from an underlying verb (páriṣṭiḥ ‘enclosing’
< pári √as ‘enclose’), with the object (Agni, an underlying objective genitive) to be
supplied from context (“the enclosing [of Agni]”). The simile contains both subject
and object in the appropriate cases (nom. dyaúḥ and [presumably] acc. bhū́ma) but no
verb, which must be generated from the noun páriṣṭiḥ (“as heaven [encloses] the
earth”), which in turn was generated from the underlying verb. Thus, the frame and
the simile are in syntactic complementary distribution – the former with a verb (or
verb substitute) but no overt nominal complements, the latter with nominal
complements but no overt verb. That the verb-y half is expressed nominally and the
noun-y half as if it contained a governing verb is a further bit of cleverness. This
syntactic skew produces case mismatch, with the nom. “heaven” of the simile
apparently compared to nom. “the enclosing” (páriṣṭir dyaúr ná), and, further, the
mismatch between the single (presumed) complement of the frame and the two overt
complements in the simile gives an impression almost of case disharmony of the type
treated in Jamison 1982, although in fact the objective gen. “of Agni” that must be
supplied matches the acc. object bhū́ma in the simile.
Or at least it matches bhū́ma acdg. to the standard renderings (Ge, Re, WG,
Lü, Varuṇa p. 590, and the publ. tr.). However, bhū́ma is a neuter and ambiguous as
to case, and heaven enclosing the earth does not strike me as a standard Vedic image
– though it slips easily into our round-earth cosmological worldview. Far more
commonly in Vedic heaven and earth together encompass something else (e.g.,
X.27.7, under an epithet) or fail to (II.16.3, again under an epithet), or they
themselves are encompassed (see the expressions quoted in Jamison 1998). For a
nicely balanced passage in which the two appear both as object and subject, see
nearby (but attributed to a different poet) I.61.8cd pári dyā́vāprthivī́ jabhra urvī́,
nā́sya té mahimā́nam pári ṣṭaḥ “He held encircled broad heaven and earth; they did
not encompass his greatness.” I therefore now wonder if bhū́ma should be taken as a
nominative parallel to dyaúḥ, with an object to be supplied (“… as heaven [and] earth
95
[encompass X]”) or even if dyaúḥ … bhū́ma should be recombined into a dual
dvandva (cf. dyā́vā-bhūmī́, with slightly different 2nd member) and taken as an object
(“as [X encompasses] heaven [and] earth”). (Note that the disyllabic reading of
diyaúḥ could accommodate a dyā́vā.) In any case, I believe that the poet kept these
possibilities open by the ambiguous concision of his wording.
Yet another aspect of this little expression is its relation to other such
expressions in this small group of hymns attributed to Parāśara Śāktya (I.65-73):
twice elsewhere in this collection Agni is depicted as encompassing something else,
with the lexemes pári √bhū (I.68.2 pári … bhúvat, I.69.2 pári … babhūtha), while in
I.72.2 something else is enclosing Agni, configured as a calf (vátsaṃ), using pári √as.
(I wonder if the somewhat pleonastic presence of bhúvat in our passage is meant to
resonate with those pári √bhū passages.) Thus, though in our passage Agni is clearly
the object of enclosure in the myth of his hiding in the waters, he himself can take the
role of encloser in this tight-knit collection – allowing for the possibility that he
might here be supplied as subjective genitive with páriṣṭiḥ, save for the mythological
context.
All in all, I.65.3cd packs a lot into its ten syllables and five words, as
witnessed by the many more words (more than ten times as many) in this entry
required to describe its syntactic and semantic tricks.
I.65.4: It is unclear with what part of the sentence the instr. panvā́ should be
construed. I take it with the verb várdhanti (though I do not go as far as Old, who
suggests that the phrase is equivalent to panáyanti), while Ge (/WG) and Re, in
slightly different ways, take it with súśiśvim. The stem is a hapax and there is no
obvious way to decide the matter, though the passages Re adduces for his instr. of
characterization (IX.85.11, 86.31 śíśum .. pánipnatam and III.1.13 apā́ṃ gárbham …
pániṣṭhaṃ jātám) may tip the scales in his direction. So I might modify the publ. tr.
to “The waters strengthen the lovely child in his wonder” or (see below) “…
strengthen the one growing well with/in wonder,” though construing it with
várdhanti still seems to yield more sense.
The matter is made more difficult by the fact that súśiśvi- is also a hapax. It is
generally taken as an adj. “schön wachsend” (ultimately to √śū ‘swell’), and this is
certainly possible. But because of śíśvā, instr. of śíśu- ‘child’, in 10c, I follow Old
(SBE, “the fine child”) and Re (“le beau Nouveau-né”) in taking it as a noun.
I.65.5: It is tempting to take the adjectives raṇvā́, pṛthvī́, and śambhú as referring to
Agni and specifying the term of comparison with the item in the simile (“broad like a
place of peaceful dwelling,” etc.), and in fact Ge (mostly) and Re (entirely) give in to
this temptation. However, all three adjectives agree in gender with the noun in the
simile (fem. in the first two, neut. in the last), and although it would be possible to
explain this agreement as “attraction” (so Re), a simpler solution is to assume that the
adjective belongs with the comparandum. (So also WG.) The phrase girír ná bhújma
poses a different problem, in that girí- is masc., and therefore if bhujmá is an
adjective, it does not modify giríḥ -- unlike the otherwise entirely parallel expression
96
in VIII.50.2 (Vālakh.) girír ná bhujmā́. We should either emend the text (à bhujmā́,
so Gr) to agree with that passage (“like a beneficial mountain”) or take it as a
nominalized neut. “source of benefit,” the solution adopted in the publ. tr. On this
characteristic of mountains, see, e.g., I.55.3.
I.65.6–7: The similes continue in these verses, but with less strict parallelism in
structure.
I.65.6: The construction of síndhur ná kṣódaḥ is clarified by its occurrences in the
next hymn, I.66.10 síndhur ná kṣódaḥ prá nī́cīr ainot “Like a river its gush, he has
sent forth those [=butter offerings?] heading downward.” Hence the “(sending)”
supplied here.
I.65.7: íbhya- is found only here in the RV, though it occurs later. It is a derivative of
íbha-, which is slightly better attested in the RV and seems to mean ‘(group of)
servants, retinue’. Tr. of íbhyān split dramatically between ‘wealthy’ (Ge “die
Reichen,” WG “die Begüterten”) and ‘vassal, servant’ (Re “les vassaux”), and the
Pāli cognate ibbha- is also glossed with both, though with the ‘vassal’ sense first and
dominant (see the newest ed. of the PTS dictionary, 2001 s.v.: “a member of a king’s
entourage; a vassal; dependent; wealthy”). I will not enter into a reexamination of all
the Sanskrit and MIA evidence, but in my opinion context here favors ‘vassal,
dependent’ rather than ‘wealthy’. See Proferes’s disc. of vss. 7–9 (2007: 111–12);
although he refuses to tr. íbhyān, he argues that it refers “to those who are
subordinated to the king,” quite possibly the clans – as the AV passage he adduces
suggests: AV IV.22.7 víśo addhi “eat the clans,” addressed to the king at his
consecration.
I.65.9: Strictly speaking, given the position of ná, the apsú should go with the frame,
not the simile: “he hisses in the waters, like …” This is presumably a reference both
to the myth of Agni’s hiding in the waters treated earlier in the hymn and to his
identification with Apām Napāt, but the haṃsá- is surely sitting in the waters, too –
though the primary point of comparison is the hissing noise both make.
How to construe gen. viśā́m is not clear. I take it loosely as a datival gen. with
the cétiṣṭhaḥ, while Re considers it dependent on a “virtual” viśpáti- (“le plus
remarquable … de (ceux qui président aux) tribus”), and Ge and WG, in slightly
different ways, take it as dependent on uṣarbhúd (Ge “der Frühwache der
Ansiedelungen”; WG “der von Niederlassungen am Morgen wach wird”). Proferes’s
rendering is quite awkward, but closest to my own in intent: “the clans' brightest as
concerns strategems.”
I.65.10: The simile in cd is unclear; its purport depends on what meaning is assigned
to multivalent vibhú-, whose senses include ‘conspicuous’, ‘distinguished’, and
‘wide-ranging’, to which Ge adds ‘sich mehrend’, Re ‘abondant’, and WG ‘kräftig’.
In Ge’s and Re’s readings, the point of the simile would be the size of the herd and
97
its growth through the birth of its young. This is certainly possible, and the point of
contact with Agni would be the growth, that is, the blazing up, of the fire after its
kindling. However, I do not see ‘increase’ as one of the core meanings of ví √bhū
and prefer ‘wide-ranging’ here, with the simile expressing the grazing behavior of
herds and the frame the fire’s tendency to spread. See 8b vánā vy ásthāt “he has
spread out through the wood.” The final word of 10, dūrébhāḥ ‘far-radiant’, may
support this interpretation. I do admit, however, that śíśvā ‘with their young’ makes
less sense than in the Ge/Re interpretation.
Note that the opening of the last hemistich of the hymn, paśúr ná śī́śvā,
“unpacks” the hymn’s first word paśvā́, and that this phrase was in a sense
anticipated by 4b panvā́ súśiśvim.
I.66 Agni
I.66.1: The fem. gender of citrā́ is a problem. The standard solution (see, e.g., Gr) is
to assume that rayí-, which is normally masc., is occasionally fem. (so also,
supposedly, in nearby I.68.7, though see disc. there). I find this unsatisfactory and
explain the gender by assuming that citrā́ modifies not only rayíḥ but also fem.
saṃdṛ́k and agrees with the latter.
Gr takes sū́raḥ as the nom. sg. of the thematic stem sū́ra-; in this he is
apparently followed by Re: “(propre à la) vision (des hommes) comme le soleil.” I
find this unlikely, since saṃdṛ́ś- regularly takes the gen., incl. sū́ryasya (II.33.1,
X.37.6, X.59.5). I therefore take sū́raḥ as gen. of svàr-, with, apparently, Ge (/WG).
More problematic is ā́yur ná prāṇáḥ, where the first word should be the neut.
noun ā́yus- ‘lifetime’, since Gr’s supposed adjectival stem ā́yu- ‘beweglich’ does not
exist (only the neut. noun ā́yu-, also ‘lifetime’). Thus, properly speaking, we should
have an equational simile “breath like lifetime,” represented in Re’s “souffle-animé
comme la vie (même)” and WG “der Atem gleichsam als Leben.” Nonetheless I
consider the interpretation as an improper compound, as in Ge’s “wie der
Lebenshauch,” to make more sense.
I.66.2: Both tákvan- and bhū́rṇi- have received diverse tr., which I will not treat in
detail here. For √tak as ‘plunge, rush,’ etc., see EWA s.v. TAK; here and in takva-vī́I take it to refer to a “swooping” bird of prey. As for bhū́rni, it belongs to √bhur
‘quiver’, and I tr. it here as ‘restless’ to capture the constant circling and diving of
birds of prey. How then can he “cling” (siṣakti) to the wood? The image must be of
the flames of the fire constantly licking and dancing (to use our preferred metaphors),
but always in contact with the wood. My “clings” is a more vivid version of the
literal “accompanies, keeps company with.” To avoid the appearance of contradiction,
the phrase might better have been rendered “ever restless, like a swooping (bird), he
(yet) clings to the wood.”
Note that the second syllable of vánā somewhat echoes the ná simile markers
of previous pādas.
98
The simile of c is syntactically problematic, consisting of a neut. NA sg.
páyaḥ ‘milk’ and a fem. nom. sg. dhenúḥ ‘milk-cow’, with the relation between them
unspecified. Ge simply takes it as a functional compound (“wie die Kuhmilch”),
though unlike “life-breath” in 1c, the order of the elements would be wrong; Re takes
dhenúḥ as if it were gen. (“comme le lait (de) la vache-laitière”), while Old (SBE)
and WG render it as two similes (Old “like milk, like a milch-cow”). I have supplied
a verb “(yielding),” allowing páyaḥ to serve as acc. object, on the basis of IV.57.2
dhenúr iva páyaḥ ... dhukṣva. None of these solutions is immediately satisfying. One
clue might be whether the adjectives in d (śúcir vibhā́vā ‘blazing/gleaming and ‘farradiant’) are applicable to either the one or the other of the nouns in the simile, but
this doesn’t turn out to be very helpful. Although both adjectives are regularly used
of Agni, they are characteristic of neither milk nor cows. One exception is V.1.3
śúcir aṅkte śúcibhir góbhir agníḥ “blazing Agni is anointed with blazing cows,” but
there the “cows” stand for products of the cow, either milk (so the publ. tr.) or, more
likely, ghee, so that a metaphorical application to either noun is possible. And it may
be that the simile in c is to be taken by itself, detached from the adjectives in d; see,
e.g., 3cd, where the two pādas, one a simile, the other not, are semantically
independent.
I.66.3: As Re notes, we might expect suffix-accented jetā́ with gen. jánānām. WG,
following Tichy (1995: 343–44) interpret the gen. in datival function (“als Sieger für
die Menschen”) to avoid nominal rection with a root-accented agent noun. But since
this accentual “rule” is often broken, I see no problem with the standard interpr. as
objective gen. “conquerer of peoples.”
I.66.5: The two halves of this verse are conceptually more complex than those
preceding. Most tr. take a and b as implicitly contrastive: Agni’s blaze is
undomesticated but he is nonetheless intimately familiar (e.g., Ge “Dessen Flamme
schwer heimisch wird, (und doch ist er) eingeboren wie die Einsicht”). This is
certainly possible. However, although I recognize the contrast between duróka- and
nítya-, I think the poet is equating the two elements in a clever play: Agni’s flame
can’t be controlled and “domesticated” anymore than a man in the grip of his own
will.
As for the second half, I find it hard to believe that the poet is implying that
one’s wife is at the disposition of everybody, with the distasteful sexual connotations
this implies -- as in Re’s “(se mettent) à disposition pour chacun comme l’épouse sur
la couche.” Rather I think this is the Papageno situation: all it takes for anyone to be
happy is his own wife in his lap.
I.66.6: The “horse” generally supplied with śvetáḥ ‘white’ is based on the regular cooccurrence of that adj. with áśva- and other expressions for ‘horse’, including
metaphorically of Agni (e.g., V.1.4). It of course also harmonizes nicely with the
chariot of cd.
99
I.66.7: In the RV it’s necessary to allow sénā- to mean both ‘army’ and ‘weapon’;
although the latter could make sense here (so Re), an army is more likely to produce
an onslaught than a single weapon.
I.66.8: As discussed in the publ. intro., this verse is quite opaque and breaks the
pattern of Agni-describing similes set up to now in this hymn (and the last one). It
has, not surprisingly, stimulated much discussion and some fanciful interpretations,
all the details of which I will not rehearse here. The first question that must be
addressed is what the first word (and the fourth), yamáḥ, represents here. Is it the PN
Yama, naming the first man and the king of the dead? or the common noun ‘twin’?
or even the common noun ‘controller’ (though we might then expect the accent
yáma-)? What one decides about this question determines the direction of
interpretation of the rest of the verse. Ge (/WG) take it as Yama, here identified with
Agni; Re and Old (and I) as ‘twin’, with the two twins displaying different
characteristics. The rhetorical structure of the verse favors this interpretation; not
only do the first two pādas begin yamáḥ, which invites (but does not require) a “the
one … the other …” reading, but the contrastive characterizations found in cd also
support it.
The next problem is the gender discrepancy between jātáḥ (m.) and jánitvam
(n.) in what appear to be parallel equational clauses. To solve this Ge (/WG) simply
supply a verb with the second, to which jánitvam serves as object: “als Yama
(erzeugt er) das künftiges Geschlecht.” But this would disturb the balanced structure
of the verse, and it seems best to allow jánitvam to be predicated of yamáḥ, as Old
and Re (and I) do. Where I differ from these latter is in my interpretation of the two
equations. Old and Re take both twins to be identified with Agni, with each “twin”
representing one aspect of Agni’s nature. I think that each twin is identified first with
another entity, which is then secondarily identified with Agni. Although this might
seem complex, it allows the other parts of the verse to bear more meaning than the
exclusive “Agni = one twin, Agni = other twin” interpretation admits. As indicated in
the publ. intro., I take the verse as reflecting the circumstances of the early morning
ritual, when the fire is kindled at sunrise and the soma is pressed for the first offering.
Under that interpretation the twin that has (already) been born (jātáḥ) would be the
sun, on whose appearance the ceremony gets started. The twin that is the substance to
be born (jánitvam) would be the soma. The two are not only twins of each other, but
each is the twin of Agni. Soma and the sun are not infrequently identified (see, e.g.,
IX.66.18 tváṃ soma sū́ra éṣaḥ “you, Soma, are the sun”); the identification of Agni
and the sun is a commonplace; and Agni can also be identified with Soma (see, e.g.,
the tṛca IX.66.19–21), and they frequently share epithets (like śúci-, discussed above
2d, though in another context).
I then take the two pādas cd as referring to the sun and the soma respectively.
The sun is “the lover of girls” because he appears with the Dawns. For the same
phrase, in the acc., referring (most probably) to the sun, see I.152.4. Whereas in
IX.86.32 our same phrase pátir jánīnām is applied to Soma, where the
“women/wives” are the milk-drinks with which he is mixed. As for the secondary
100
reference to Agni, of these characterizations, either group of females (or both) might
refer to the offering ladles (fem. juhū́-) that approach the ritual fire with their
oblations or the streams (fem. dhā́rā-) of melted butter being offered. See the fem. pl.
nī́cīḥ in 10b.
Despite the complications of this interpretation, it provides a rich reading of
the enigmatic verse, which fits well with the two ritually centered verses that end the
hymn.
I.66.9: As indicated in the publ. intro., I consider the last two verses of the hymn to
reflect the ritual situation. This scenario is supported by the presence of 1st and 2nd
pronouns (vayám and vaḥ respectively), with the 2nd ps. enclitic indicating, as often,
the other human participants and observers besides the poet and priests, who are the
“we.”
The crux in this verse is carā́thā. The first problem with it is the long ā in the
nd
2 syllable, the only instance of this stem beside regularly formed carátha- (13x); 4
of these occur in the Parāśara hymns (I.68.1, 70.3, 7, 72.6), but of these, 3 (68.1, 70.3,
7) would be metrically better if read *carā́tha-. I have no explanation for the
variation and will not pursue the issue further. More interesting is the grammatical
identity of the form and the semantic role it plays in the verse. Old (SBE, Noten) and
Ge (/WG) take it as a nom. pl. ‘wanderers’, neuter if flg. the Pp., though Ge suggests
that it might reflect masc. carā́thāḥ, and interpret it as conjoined with vayám (Old,
Ge) or identified with it (WG). However, in the Noten (ad loc., fn. 1) Old allows the
possibility that the form could be an instr. parallel with vasatyā́ (“mit Gehen und mit
Verweilen”), an idea that Re develops, suggesting that the pair is semantically
parallel to the contrastive pair yóga / kṣéma (roughly “activity and rest”). Re believes
that the instrumentals characterize Agni (“(soit) dans (sa) marche, (soit) dans (sa)
demeure”). This is possible, though it would be a slightly odd use of the instr. With
Re I consider carā́thā an instr. contrasting with vasatyā́, but think that the pair is
applicable to “us”: we approach Agni with homage with both our movable goods,
that is, the livestock that provides the butter and milk offerings, and our household
establishment that supplies the rest.
The 3rd pl. nákṣante (nákṣanta in sandhi) does not agree with its 1st pl. subject
vayám. With Ge I take it as attraction to the immediately preceding simile, ástaṃ ná
gā́vaḥ. It is accented because it begins the pāda.
I.66.10: The fem. pl. nī́cīḥ has almost too many possible referents -- waters, flames,
cows, or butter offerings (configured as cows) – all of which have been proposed by
one or more tr. I favor butter offerings. See also VIII.101.13, also containing nī́cī and
showing the same range of possibilities; sim. V.44.4. The cows of c may well be the
same offerings metaphorically.
See intro. for the ring compositional final pāda. As Re points out, svàr dṛ́śīke
is a nominalizing extension of the infinitive phrase svàr dr̥śé “to see the sun,” and we
need not try to interpret svàr as a genitive.
101
I.67 Agni
I.67.1: ‘Demand’ may push √vṛ ‘choose’ a bit, but the simile suggests a degree of
coercion.
Adverbial ajuryám is lit. ‘undecaying, unaging’, but in the developed sense
‘unfailing’ or, in English idiomatic terms, ‘without fail’.
I.67.3: On the idiom áme √dhā see disc. ad I.63.1; this expression must also be
considered in relation to Parāśara’s ámaṃ √dhā in the preceding hymn (I.67.7).
Although a tr. like WG “setzt er die Götter in Schrecken” is easier, I do not think we
should lose the sense of ‘onslaught’ for áma- by redefining it as ‘panic’ or ‘terror’ in
this idiom. Here I think the point is that the leaping flames of Agni put even the gods
in his way, even though he is, for the moment, sitting undetected. The first pāda of
the hymn (váneṣu jāyúḥ “a conquerer amid the wood”) also implicitly contrasts his
martial power with his apparently humble station.
I.67.4: The anomalous 3rd pl. vidánti, to the thematic aorist ávidat, is the only
apparent primary indicative form to this stem and should belong to a present.
However, I think it can be eliminated: rather than following the Pp. vidánti īm, we
can instead read vidánt īm, an injunctive with secondary ending whose final -t was
preserved in this ambiguous sandhi situation. No alteration of the text is required.
The publ. tr. need not be altered either, though a preterital interpretation might be
better: “They found him here when they recited mantras…” – esp. if this refers to the
mythological tale of the gods finding Agni hidden in the waters. Cf. I.72.2 (also a
Parāśara product), where the immortals did not find him (ná vindan) though they
sought him; they are also described as dhiyaṃdhā-, as here.
I.67.5: Tr. of the subj. ajáḥ vary between ‘unborn’ (Ge, Re) and ‘goat’ (Old SBE,
WG), but parallel passages show that it must be the shadowy divinity Aja Ekapad
(“One-footed Aja), whatever the ajá- signifies in that conjunction. (Old recognizes as
much in his n.) Cf. esp. AV XIII.1.6 tátra śiśriye 'já ékapādó 'dr̥ṃhad dyā́vāpr̥thivī
bálena; also X.65.13 where ékapād ajaḥ is adjacent to divó dhartā́. In our passage the
word padá- appears in the next verse and gestures toward the epithet.
The pf. dādhā́ra is ordinarily always presential (Kü 261); Kü cannot decide
whether the pf. of √stambh is presential or preterital, but at least in this context I
think it matches dadhā́ra in value.
I.67.6: The verse is highly alliterative: ab: priyā́ padā́ni, paśvó ní pāhi; d: guhā́
guháṃ gāḥ. For the stylistic figure guhā́ guhám see Hoffmann, KZ 76 (1960) 242–
48, esp. 246–47 [=Aufs. I.113-19]. Note that guhā́ has a different accentuation from
adverbial gúhā.
I.67.7–8: These two verses are syntactically dependent on each other, consisting of
three relative clauses (7ab, cd, 8ab) and a main clause with coreferential pronoun
102
(8cd). However, this structure is complicated by the fact that, though two of the rel.
clauses have singular rel. pronouns (yáḥ 71, 7c) and the pronoun in the main cl. is
likewise singular (asmai), the intermediate rel. cl. (8ab) has pl. yé, though all seem to
be roughly coreferential. There is no obvious solution for this jarring inconsistency;
Ge simply labels it Anakoluthon. However, certain features of the larger context may
have encouraged this syntactic anomaly. First, the relative/correlative construction is
of the indefinite “whoever … “ type, where number is functionally, though not
grammatically, neutralized. That is, it doesn't matter how many (or few) people the
definitional relative clauses identify: if many people perform the functions, they
should all get the reward; if only one does, then only he would. Moreover, as Lüders
argues (p. 448), vss. 7–8 respond to vss. 3–4. In those earlier verses Agni goes into
hiding (3d), but the “superior men” (náraḥ) find him after they produce and recite
effective speech (4). In vss. 7–8 we seem to have the same situation, but in the
human realm: he/they who has/have recourse to and do service to the truth (7cd, 8ab)
and perceive(s) Agni in hiding (7ab) receive(s) acknowledgement from Agni himself
(8cd). Since in vss. 3–4 the discoverers of Agni, the gods, are plural, it may be that
the plural crept in here, too. And finally this sequence may anticipate the next hymn,
where it is emphasized that “all” (in the plural) perform the proper ritual and
devotional functions, and in particular the almost identical phrase ṛtáṃ sápantaḥ
(68.4c) echoes our ṛtā́ sápantaḥ (67.8b).
I.67.8: The verb ví … cṛntánti lacks an overt object, and several different ones have
been proposed: Ge (in n.) suggests it’s “the stream of truth” from 7d; Re supplies
“(l’énigme)”; and WG read ṛtā́ with this verb as well as the participle sápantaḥ
(“welche … die Ṛtas aufknoten, (sie) pflegend”). Since the other two occurrences of
this root (impv. ví … cṛta I.25.21, ppl. vícṛtta-) are both construed with pā́śa- ‘fetter’,
that seems the likely object here as well. More speculatively, on the basis of VII.59.8
druháḥ pā́śān ‘fetters of deceit’, I have supplied that whole phrase here, since ‘deceit’
contrasts nicely with the emphasis on ṛtá- ‘truth’.
The pf. prá vavāca is one of the very few pf. forms to √vac with full
reduplication (against uvā́ca, etc.). On the basis of its agreement with Aves. vauuac-,
Kü (p. 441) considers this the older type of reduplication, with the newer type
spreading from Samprasāraṇa roots.
I.67.9: The second half of the verse is metrically disturbed. As HvN note in their
metrical commentary, it actually reads better as a Triṣṭubh, as is occasionally the
case with Dvipadā Virāj.
The syntax of the verse is also problematic. The two utá of c are difficult to
construe, and partly for this reason Gr, Old (SBE [disavowed in Noten]), Re, and
Klein (DGRV 358) emend prajā́ utá to prajā́su (presumably then *prajā́sūtá in the
Saṃhitā text, though it’s not stated). If the original text had read this, it is hard to see
why it would have become corrupted, given the two other loc. pl. in this verse. I
consider the problematic utás and the problematic prajā́ḥ connected and suggest that
the utás are connecting different entities: the first connects the nominatives yáḥ …
103
utá prajā́ḥ “who [Agni] and his offspring” (not grammatical in English, of course,
but possibly so in Skt; consider the common “X and which Y” construction), the
second the locatives vīrútsu … utá prasū́ṣu antáḥ “in the sprouts and within the
fruitful (plants).” Both Agni and his offspring grow in both media; note that because
of its sandhi position ródhan can be both sg. (ródhat) and pl. and thus construed with
both suggested subjects.
Note also that the verse begins ví REL, just like vs. 8, but with entirely
different referents for the rel. pronoun. This might be another reason why 8ab shifts
to the plural, to establish the difference between those who serve Agni and Agni
himself.
I.67.10: The exact sense of cittí- and the syntactic disposition of apā́m are the two
issues in this verse. Although cittí- ordinarily means ‘perception, insight’ and most tr.
so take it, I prefer to make it the abstract to a different, but well-represented, sense of
√cit, namely ‘appears’, hence ‘apparition’. I also, with Old SBE, take apā́m with
dáme rather than cittíḥ. I find tr. like “la pensée active des eaux” (Re), “der Verstand
der Wasser” (WG) hard to interpret, even lacking sense. Since much of the hymn has
dealt with the theme of Agni’s hiding – and since his hiding place was the waters,
though this was not overtly stated earlier – I think that this verse describes him
glowing in the waters (and therefore detectible) and also makes reference to him as
Apām Napāt.
The sádma of c must be read with both simile and frame.
I.68–70 Agni
I consider these three hymns to be thematically and verbally linked, treating
the Vaiśvānara fire and encoding this theme by deploying the two words underlying
this vṛddhi compound, víśva- and nṛ́-. The first, víśva-, dominates I.68, the second,
nṛ́- I.69, and they are juxtaposed in the climactic verses of I.70.9–10. See the
introductions to the three hymns in the publ. tr., and for a detailed discussion,
Jamison forthcoming (“Inter-hymnic Rhetorical Structure in Ṛgveda I.68-70:
Parāśara Śāktya’s Vaiśvānara Cycle”).
I.68 Agni
Note that Hoffmann tr. the entire hymn (1967: 141–42) because it contains 10
injunctives. I agree with Hoffmann’s tr. of these forms as presents, although I do not
think this requires a Zeitlos-type interpretation.
I.68.1: For this interpr. of śrīṇán, which lacks overt object, see Narten 1987: 281
(=KlSch p. 351).
Rather than taking aktū́n as a second object with vy ū̀rṇot, I make it an acc. of
extent of time, as it often is elsewhere (e.g., V.54.4).
On the metrical shape of carátham see I.66.9.
104
I.68.2: The second half-verse does not work as Dvipadā Virāj, since it would have
pādas of 4 and 6 syllables. It has long been suggested (for reff. see Old SBE and
HvN comm., both ad loc.) that putting devó devā́nām in the opposite order would fix
this problem. However, Old argues in the Proleg. (97) that this is unnecessary, that
Dvipadā Virāj has an affinity to Triṣṭubh, and that this line, though 10 syllables,
configures itself nearly as a Triṣṭubh. See I.67.9 above.
The verb pári … bhuvat lacks overt object, and various suggestions have been
made. I assume that it is the object of 1cd “the still and the moving.” Hoffmann (sim.
WG) tr. the verb simply as “überragt” (stands out), but this interpr. ignores the
emphasis on “encompassing” in this set of hymn (see disc. ad I.65.3).
The unaccented gen. pl. eṣām presents a problem similar to that posed by
āsām in vs. 7, though I account for them in different ways. Given its lack of accent,
eṣām should be a pronominal, not adjectival, demonstrative, but it appears to be
construed with víśveṣām … devā́nām “of all these gods.” Because devā́nām is fairly
distant from both eṣām and víśveṣām and appears to be bound to immediately
preceding deváḥ, I would now take eṣām … víśveṣām together, separate from
ddevā́nām and emend the tr. to “When alone of all these, the god among the gods,
encompasses …” Though it may be difficult to maintain this in the face of vs. 7, as I
argue in the forthcoming article cited above, the poet is doling out the genitive
plurals here and does not want to specify that “these all” are the gods until the last
minute, since masc. pl. “all” in this hymn otherwise refers to the unified worshipers
of the Vaiśvānara fire.
I.68.4: The publ. tr. has “All have a share in divinity and its name.” I would now
replace this with “… your divinity and your name.” As I argue in the art. cit., the “all”
literally share the name Vaiśvānara.
The contrast between joint enterprise of those unified by the Vaiśvānara fire,
expressed by bhájanta víśve, and the separate actions of each of the members,
expressed by évaiḥ, is the theme of this verse as well as vs. 8.
I.68.5: The logical and syntactic connections between the nominal expressions of ab
and the clause in cd are not clear. Ge simply pronounces ab elliptical and suggests
supplying cakruḥ from cd, while Re supplies “(telles sont/est)” with the two phrases.
My assumption is that the promptings and vision of truth are the indirect cause of the
actions in cd by defining the proper tasks that all should perform throughout their
lifetimes.
I.68.7: There is a syntactic problem in cd that is ignored by all tr. as far as I can see,
though Old treats it briefly (ZDMG 61 [1907]: 829 = KlSch p. 260; see below): the
fem. gen. pl. demonstr. āsām is unaccented. Oblique forms of the ayám
demonstrative are unaccented when used pronominally, but accented when used
adjectivally. The two accented forms of āsā́m appear with NPs, pū́rvāsām …
svásṝṇām “previous sisters” I.124.9 and viśā́m .. ábhayānām “fearless clans” X.92.14,
and unaccented āsām (25x) is always pronominal (for VII.34.10 and X.75.4 see disc.
105
ad loc.) Yet all interpr. of our passage construe āsām with rayīṇā́m (e.g., Ge [/WG]
“der Herr dieser Reichtümer,” Re “le maître de ces richesses”). This produces
another anomaly: rayí- is overwhelmingly masculine and should not be modified by
a feminine demonstrative. (On supposed fem. rayí- in I.66.1 belonging to this group
of hymns, see disc. ad loc. It need not, and in my opinion is not, feminine there
either.) Although it introduces some complexity, I therefore think that āsām “of these”
must stand for a different feminine noun also construed with páti-. The problem then
is what noun? It should already be present in the discourse, since unaccented forms
of ayám are anaphoric, but there are no obvious candidates – indeed, no candidates at
all if we limit ourselves to overt feminine plurals in the previous verses of this hymn.
However, an underlying feminine referent can be generated from the context. I see
two possibilities. 1) From ápatya- ‘progeny’ in the phrase mánor ápatye we can
generate the synonym, fem. prajā́- ‘offspring’ – note their juxtaposition in I.179.6
prajā́m ápatyam. The gen. pl. of prajā́- does not occur in Vedic (indeed no gen. pl. to
a root noun in -ā does, see Macdonell VG), presumably because it should be prajā́m
and identical to the acc. sg. (though with possible distracted 2nd syllable). By this
interpretation Agni would essentially be prajā́-pati. 2) More likely, in my view. The
underlying noun is víś- ‘clan’. Although no direct reference to clans is made in the
hymn, as I point out in the publ. intro. and argue in more detail elsewhere (Jamison
forthcoming), the repeated víśve ‘all’ triggers a pun with víś- in this set of hymns (see
esp. I.70.4) and so would be present in the minds of the poet and his audience. Agni
is regularly called viśā́m páti- and viśpáti-, sometimes at the same time – e.g.,
III.13.5 hótāraṃ viśpátiṃ viśā́m, which also contains hótar-, as also here. The
mention of Manu might also have triggered the association; cf., e.g., V.4.3 viśā́ṃ
kavíṃ viśpátim mā́nuṣīnām “the poet of the clans, the clanlord of the (clans) of the
sons of Manu.” If this solution seems too fussy, it is possible to follow Old’s
explanation of the lack of accentuation (see ref. above): that āsām does not modify
rayīṇā́m but anticipates it (“he is the master of them, of the riches”), while being
anaphoric to sg. rayím in 6d. This does not eliminate the gender issue, however, and
also seems over-tricky. The eṣām … víśveṣām … devā́nām of vs. 2 makes some
problems for my interpretation here, but see discussion there.
I.68.8: This verse has given rise to a number of different interpretations, which
cannot be discussed in detail here. I interpret it in the framework of Proferes’s
treatment of the joint clan-fire (see publ. intro.), which must be kindled and tended
by members of the separate clans working together. This cooperative and reciprocal
enterprise is expressed by the phrase mithás tanū́ṣu “mutually among themselves” (at
least in my interpretation; it has received varying tr., but with most rendering tanū́ṣu
as ‘bodies’), sáṃ jānata “they agree, act in unison” to the resonant lexeme sám √jñā
(see publ. intro. to I.68 and esp. I.68.9), and by the middle voice of ichanta “they
seek (from each other).” The rétaḥ they seek is, I think, the means of kindling the fire,
though I admit that this is not a usual sense of that word. Again, as in 4cd, there is a
contrast between joint action (sáṃ jānata) and each individual’s contribution to it
(svaír dákṣaiḥ).
106
I.68.9: This verse returns us to vs. 3; the b-pādas of both are identical: krátuṃ juṣanta
“They take pleasure in your/his resolve.” In vs. 3 the subject is “all”; here we can
assume that the unidentified subject here is “all” as well, and, since vs. 8 follows
easily on vs. 7, we can superimpose “all” as subject there too. In vs. 3 the “all” found
this pleasure when Agni was born in cd; the parallelism between vs. 3 and vss. 8–9
supports my view that vs. 8 also concerns the kindling of the fire, and the birth
metaphor of 3cd (jániṣṭḥāḥ “you are born”) further supports my suggestion that the
rétaḥ ‘seed’ of 8a refers to the means of begetting the fire.
I.68.10: As Hoffmann points out, ví … aurṇot is the only augmented form in the
hymn and is esp. striking because it forms a ring with vy ū̀rṇot in 1d. The use of a
clear past tense form seems to me to mark the conclusion of the ritual kindling of
Agni and the attendant distribution of largesse. In other words, like so many final
verses it summarizes the ritual activity whose general description precedes it. The
final half-verse, in which Agni decorates the vault with stars, may seem out of place,
but it also forms a ring, thematic not lexical, with vs. 1, where Agni approaches
heaven and performs his tasks through the nights: the vault (nā́kam) reprises heaven
in 1ab, and the stars are appropriate to the nights.
Strictly speaking, rā́yaḥ is acc. pl., not gen. sg. (rāyáḥ), and a more literal
translation would follow Ge’s “Er schloss die Reichtümer, die Töre (dazu), auf,” or –
to match my tr. of 1cd and accommodate English idiom – “he disclosed riches,
opened wide (its) doors.” However, Old strenuously argues that it should be
interpreted as gen., and certainly the idiom rāyó dúraḥ (with the correct accent) is
found nearby in Parāśara’s oeuvre (I.72.8)
What exactly is meant by Agni’s decorating heaven with stars is unclear.
However, in a similar passage (IV.7.3) Proferes (p. 27) argues that the stars in the
sky represent the domestic fires distributed though many households.
I.69 Agni
As discussed in the publ. intro. and in the art. cit., this hymn is twinned with I.68.
I.69.1: The etymological figure śukráḥ śuśukvā́n is not as redundant as it seems,
because both forms are found in similes with the sun: svàr ṇá śukrám (II.2.7,
IV.45.2), with the adjective, and svàr ṇá śuśucīta (II.2.10, X.43.9), with the perfect,
both exist (together in X.43.9 svàr ṇá śukráṃ śuśucīta). In other words, since the
similes "like the lover of Dawn" and “like the light of heaven” have the sun as their
underlying referent, at least one of the √śuc forms is, as it were, bound to it
formulaically.
I.69.2: This verse is parallel to its twin in I.68, with the same verb: #pári …
babhūtha [/bhúvat 68.2]… ; the emphasis on Agni’s status among the gods (devó
devā́nāṃ 68.2, devā́nām pitā́ 69.2); and an instr. in -tvā (mahitvā́ 68.2, krátvā 69.2),
though to non-parallel stems (mahitvá-, krátu-). It is therefore surprising that tr. do
107
not note the parallelism of the two verses or necessarily treat them in the same way.
In particular, both Ge and Old (SBE) take the pári…bhúvat in 68.2 as ‘encompass’,
but pári … babhūtha in 69.2 as ‘be superior, excel’. (Re tr. both as ‘encompass’ and
WG both as ‘excel’.) Given the focus on encompassing/surrounding in this set of
hymns (see disc. ad I.65.3 above), I think both lexemes should be tr. the same way,
and each requires an object to be supplied. In 68.2 I supply “the still and the moving”
from the previous verse; here I would use the conjoined world-halves, also from the
previous verse.
The poet handles the parallelism between the two verses in almost a
syncopated fashion. As just noted, each verse contains the idiom pári √bhū; in 69.2
this is contained within the first hemistich, whereas the finite form bhúvat opens the
second hemistich in 68.2. But 69.2 has the almost identical form bhúvaḥ in that same
position, but not as part of the idiom (see also bhúvat I.67.2, 65.3). As Hoffmann
argues (e.g., 1967: 236–37), bhuvaḥ, bhuvat are formally ambiguous, both injunctive
to the thematic aorist and subjunctive to the root aorist of √bhū. In the publ. tr. I
render bhúvaḥ in I.69.2 as a subjunctive, but now I am inclined to take it as a
presential injunctive “you become…” parallel to my interpretation of bhúvat in 68.2.
I.69.4: I supply mitráḥ ‘ally’ on the basis of a number of similar formulaic phrases:
jáne mitró ná X.22.1; mitráṃ ná jáne VIII.23.8 (of Agni); mitrá iva ... jáne II.4.1 (of
Agni). This would be via the collocation mitrá śéva; cf. nearby I.58.6 mitráṃ ná
śévam divyā́ya jánmane “favorable like an ally to the divine race,” also of Agni, with
a different word for ‘people, race’ (cf. also śévam mitrā́ya X.113.5, a diff.
permutation). Note also, two hymns previously, I.67.1b márteṣu mitráḥ, where
márteṣu is a reasonable equivalent to jáne.
āhū́rya- is a hapax, universally taken to belong to √hvṛ ‘go crookedly, go
astray’; although this derivation is not impossible, it leads to some forced tr. (e.g., Ge
Durchgänger [‘bolter’, of horses]). I take it rather as the gerundive to the seṭ root √hṝ
‘be angry’.
I.69.5: The phrase víśo ví tārīt is variously interpreted. The first issue it raises is
whether to take the acc. víśaḥ as direct object (so Ge, Re, WG) or to construe it with
the preverb ví as acc. of extent of space (Old SBE). Both usages are attested for ví
√tṝ, but it is worth noting that the latter usage is found in Parāśara’ oeuvre using the
same form ví tārīt (I.73.1), and I adopt it here. In the simile, “traverse the clans” may
refer to the year-long journey of the horse ultimately sacrificed in the Aśvamedha. In
the frame it reflects the idea of the Vaiśvānara fire spread through all the clans, not
limited to a single household or small family group.
I.69.6: The intrusion of the 1st sg. speaker through áhve “I have summoned” is
remarkable. Who is he? Although ordinarily in the RV the default referent of a 1st ps.,
esp. a 1st sg., is the poet, I wonder in this instance if it is not the leader of the united
forces, the Kriegskönig, calling the clans and their leaders and best warriors to unite
them for action.
108
My tr. differs syntactically in several ways from most tr., though it agrees
with Re in both. First, I do not construe instr. nṛ́bhiḥ with sánīḷāḥ (“of the same nest
with the men”) but as an independent instr. of accompaniment. This would be the
only ex. of sánīḷa- with an instr. By specifying “the clans along with their men,” the
poet both refers to the whole social group and singles out its most conspicuous and
important members, perhaps the viśpáti-s.
Second, I take devatvā́ as an instr. sg., not acc. pl. neut. with víśvāni. There
are several reasons for this. The first verse of the next hymn (I.70.1) contains the
same phrase víśvāny aśyāḥ, without further specification, and if at all possible the
two phrases should be rendered in the same way. Moreover, as Re points out, no
direct cases of the plural are attested in this type of derivative. It’s also the case that
Agni is characterized by instrumentals in -tvā in this group of hymns (mahitvā́ I.67.9,
68.2; krátvā I.69.2), and his divinity is emphasized (I.68.2, 4, 69.2). I am, however,
given pause by several passages containing devatvám (sg.) √(n)aś: I.151.9 ná
devatvám paṇáyo nā́naśuḥ … and III.60.2 téna devatvám ṛbhavaḥ sám ānaśa.
I.69.8: The first hemistich has 11 syllables and is best interpreted as a Triṣṭubh, with
HvN.
The verse is otherwise problematic, and my interpretation differs sharply
from that of others. The crucial point of difference is áhan, which is universally
taken as the 2nd sg. impf. of √han ‘smite’. This seems to me contextually very
unlikely: √han is a very rare verb with Agni as subject, and when it is found, it is
almost always of Agni’s special form as rakṣo-hán- (see IV.3.14, V.2.10, VI.16.29,
and with a different object V.4.5). The context does not favor a verb of violence, and
I find it hard to believe that √han would be attributed to Agni out of the blue, esp.
without specifying an object and esp. announcing this action as his particular
“wondrous power” (dáṃsaḥ). Instead I take it as the loc. sg. of áhar ‘day’, in the
formula samāné áhan “on the same day” (cf. I.34.3, 186.4), with the samāné
suggested by (or suppressed by) adjacent samānaíḥ. (Something like this possibility
is considered by Old in his n. in SBE and attributed to Aufrecht in Noten.)
The major argument against this interpr, as far as I can see, is the double yád,
which is unusual under any interpretation but passes better with two verbs (yád áhan
… yád … vivéḥ) than one. I do not have a good solution for this doubling. Either it is
rhetorical or simply pleonastic, or it marks off nṛ́bhiḥ … yuktáḥ as a quasi
circumstantial clause: “This is your wondrous power, that (yád 1) on the (same) day,
when (yád 2) yoked with the same men, you …” But I know of no parallels for such
a usage. Another argument for seeing √han here, made by Old (Noten), comes from
X.147.1, which contains both áhan and vivér apáḥ (see immed. below), but I do not
think that very late hymn, dedicated to Indra, should overly influence our
interpretation of this passage.
With Old (SBE and Noten) I read vivér ápāṃsi, against Pp. vivé rápāṃsi;
Old’s parallels are very persuasive, and this analysis does not require a change in the
Saṃhitā text. Ge’s tr. appears to follow the Pp. reading, however, as does Re’s,
though in his n. he accepts Old’s reading.
109
I.69.9: The first pāda is identical to 1b and thus forms a ring.
Gr, Ge, Re (apparently), and Lub take usráḥ as gen. to uṣár-, but since there
exists a robust stem usrá- ‘ruddy’, I see no reason not to take it as nom. sg. to that
stem, sep. since vibhā́van- doesn’t otherwise take a gen.: the three passages Ge
adduces are not compelling.
As discussed in the publ. intro., I do not follow the standard interpr. of
sáṃjñāta-rūpa-, i.e., “having (his) well-known form,” since the idiom sám √jñā has a
quite specific meaning ‘come to agreement’. The idiom is distinctive enough, and the
compound is heavy enough, that that idiom is surely meant, esp. because it appears in
finite form in I.68.8. Instead of being an anodyne description of the fire, this is a
meaningful characterization of the shared clan-fire as the symbol of, and the product
of, the mutual and unified ritual activity of the clans. Unfortunately, to render this
comprehensibly in English requires a heavy and awkward locution.
The referent of asmai is not entirely apparent. The standard solution, going
back to Sāy, is that it is the sacrificer. This is basically correct. However, the lack of
accent on the pronoun should require that the referent is already present in the
discourse. I suggest that it is to be found in the 1st ps. subject of áhve “I have
summoned” in vs. 6, a subject that may be one of the ritual participants or, as
suggested above, the Kriegskönig. Switch between persons is of course extremely
common in the RV.
I.69.10: The phrase dúro vy ṛ̀ṇvan echoes ví … aurṇod dúraḥ, which opens the last
verse of the preceding hymn (I.68.10) and which itself forms a ring with vy ū̀rṇot of
I.68.1.
The tmánā ‘by himself/themselves’ must refer to the priests; since Agni is
regularly the subject of √vah, as the conveyor of oblations to the gods and of the
gods to the sacrifice, it is here emphasized that others are conveying him.
The last hemistich is identical with the last one of I.66.10, which itself forms a
ring with I.66.1. This set of hymns is very tightly knit together! As argued in the art.
cit., this verse marks the completion of the fire’s transformation into the Vaiśvānara
fire. Since that fire is identified with the sun, “all cry out on seeing the sun” is a
expression of the unified group’s first sight of and recognition of the fire that
symbolizes their unity – which has been presented to them by the priests who carry it
and throw open the doors for it to be seen in 10ab.
I.70 Agni
Again, for detailed discussion of the thematics of this hymn and its relation to
the two that precede it, see the publ. tr. and the art. cited above.
I.70.1: It is perhaps appropriate that this last hymn in the Vaiśvānara series begins
with the martial verb vanema “may we vanquish,” since the function of the
Vaiśvānara fire is to unite the clans as a force to oppose its enemies.
110
The Pp. reads manīṣā́; most tr. read manīṣā́(ḥ) (see Old’s disc [Noten]).
However, I take it as both an instr. sg. and acc. pl., with the poet, as so often, taking
advantage of potential ambiguity. This double reading is supported by I.73.9 (also
Parāśara’s work), which contains the athem. opt. vanuyāmā corresponding to them.
vanéma here and three pairs of instr. + acc.: árvadbhiḥ … árvato nṛ́bhir nṛ́n
̄ , vīraír
vīrā́n vanuyāmā “Might we vanquish steeds with steeds, men with men, and heroes
with heroes.”
I.70.3: sthātā́m and caráthām are pseudo-genitive plurals to this merism much
favored by Parāśara. There’s no orthodox way to generate them morphologically;
Re’s suggestion that sthātā́m is a compromise between *sthātrām and something that
looks more like a gen. pl. and that caráthām simply copied it seems a reasonable
proposal, however sketchy the details.
I.70.4: A verse subject to numerous clashing interpretations. To deal with ádrau first
– Ge cites several passages in which Agni is born from or is the son of a/the stone. I
think it also likely that it is meant to contrast with duroṇé; the fire in the house is
actual and visible, while the fire in the rock is merely latent and not easily acquired.
Most tr. take the two hemistichs as independent units, which then necessitates
supplying something (what depends on the tr.) to make ab an actual clause. I prefer
to take the whole verse as a single clause, with svādhī́ḥ as the predicate, whose
complement is asmai (“… is very attentive to him [=Agni]”).
This leaves the enigmatic expression viśā́ṃ ná víśvo amṛ́taḥ. Most tr. want
this to refer to Agni; hence their interpretational separation of the two half-verses. I
think rather that it refers to those who serve Agni, both gods and humans. The two
groups are expressed in different ways and the connection between them is
actualized by a pun on víś- ‘clan’ and víśva- ‘all’; this pun in turn makes the secret
connection with Vaiśvānara, the clan fire (see Proferes p. 48). In the phrase, víśvaḥ
can be read both with amṛ́taḥ (“every immortal”) and with the gen. pl. viśā́m “every
(one) of the clans,” thus providing an indissoluble link between the mortal and divine
realms. And the ná makes clear that gods are exactly like men in their devotion to
Agni. I am reasonably convinced that this is the correct interpretation, but it does
trouble me that it is apparently identical to Max Müller’s (as reported by Old, SBE
n.), “To him also who dwells in the rock and in the house, every immortal like every
one among men is well disposed.”
I.70.5: The interpretation of kṣapā́vant- as ‘earth-protector’, containing a zero-grade
form of the archaic noun kṣám- ‘earth’, goes back at least to the Petersburg
dictionary and is the overwhelming consensus in Western scholarship (see EWA, s.v.
kṣám-). However, a connection with the root noun kṣáp- ‘night’ has a more ancient
pedigree: the Pp. divides the word kṣapā́ ‘vān, and Sāy comments kṣapeti rātrināma /
rātrimān / āgneyī vai rātriḥ. At the very least we have a potential pun, and this pun is
actually realized in this hymn, since vs. 7b, two verses later, contains the relatively
rare root noun kṣáp-. Though vs. 5 has no mention of night and ‘earth-protector’
111
works fine in that context, the surfacing of kṣáp- ‘night’ so soon after might make the
audience reconsider and produce a secondary semantic association for kṣapā́vān.
With a suggestion of Scarlatta (1999: 303), we could then analyze kṣapā́vant- as
based on a syntagm with original predicative instrumental (kṣapā́ “[he is] with
night”), which was then provided with a –vant- possessive suffix. Scarlatta (1999:
303) also suggests other ways to incorporate kṣáp-‘night’, e.g., by haplology from
*kṣápā + pā́- ‘protecting by night’ (his reconstructed initial accent reflects a posited
adverbial acccent shift from inst. kṣapā́; see p. 303 and n. 452). The exact details
matter less than the fact that the Vedic audience could likely see a pun in this word,
between kṣa- as a combining form of kṣám- ‘earth’ and kṣap- ‘night’. The publ. tr.
“protector of riches on earth” reflects the standard Western interpr., though with an
adjustment to incorporate rayīṇā́m; I do believe that the pun on ‘night’ was available
to the audience, however. (See also X.29.1 and Jamison 2015 [Gerow Fs., IJHS 19].)
It should also be noted, however, that the analysis ‘earth-protector’ for kṣapā́vān is
reinforced in the next vs. (6ab) by an apparent paraphrase etā́ … bhūmā ní pāhi
“protect these worlds.” The poet seems to be even-handedly offering alternative
views of kṣapā́vān.
Most tr. take ab as a nominal clause with suppressed dative (asmai or the like),
the antecdent of the yáḥ of the rel. cl. (“Agni is the protector … [for him] who…”). I
follow the interpr. of Velankar (1993: 41), who takes dā́śat as the verb of both main
and relative clauses, accented in the former because it is initial in its pāda (and in a hí
clause), in the latter because it’s in a subordinate clause. This grammatical and
semantic reciprocity would match that of the (asmai …) asmai, where the first,
suppressed dative refers to the worshiper and the second overt one to Agni. The
theme of reciprocity is prominent in this hymn group.
I.70.6: mártān is almost surely an irregular gen. pl., rather than the acc. pl. it appears
to be. (Though “knowing the races of gods and (knowing) mortals” is possible, the
tight formulaic connection between gods and men/mortals throughout the RV
strongly favors the gen. pl.) Whether it is an archaism or is simply following the
morphologically sketchy lead of sthātā́m … caráthām in vs. 3 cannot be determined.
I.70.7: The phrase pūrvī́ḥ kṣapó vírūpā(ḥ), lit. “many nights of differing form,” is
convincingly taken by most commentators as an elliptical pl. for “… nights (and
dawns)…,” since vírūpe regularly modifies the dual dvandva náktoṣā́sā ‘night and
dawn’, including in Parāśara’s I.73.7. (See Old’s lucid presentation in SBE n.) Most
tr. take it as the nom. pl. subject of várdhān, but, with Ge, I take it as an acc. pl.
expressing extent of time, supplying a pronominal subj. ‘they’, picking up “the races
of gods and men” in the previous vs. Either interpr. is possible, and there is little to
tip the balance one way or the other. Old argues that under the nominative interpr.,
which he favors, kṣapáḥ should probably be accented *kṣápaḥ (there are no other
nom. pl. forms), and the need to account for the wrong accent (if such it is) and the
fact that almost all occurrences of kṣapáḥ express extent of time (save for VIII.41.3,
where it’s a direct obj.) might favor the acc. interpr.
112
With essentially all Western commentators starting with Benfey, I read
carátham for the Pp. ca rátham, producing yet again Parāśara’s beloved, but
morphologically troubled, merism “the still and the moving.” With most tr. I take it
as neut. sg. and a second subject of várdhān (adjusted for number), though Ge
interprets it as a gen. dependent on the “Keim” he supplies.
The ppl. právīta- always means ‘impregnated’ (of females) or, as here,
‘conceived’ (see Scar p. 501); WG’s “den vom Ṛta gesuchten” seemingly rests on the
root etymology to √vī ‘pursue’, without taking into account the idiom.
I.70.8–9: See Proferes (pp. 47–49), esp. for the identification of the Vaiśvānara fire
in particular with the sun, and the intro. in the publ. tr. and the art. cit, esp. for the
encoding of the Vaiśvānara fire by the successive grammatical subjects víśve (9cd)
and náraḥ (10ab).
I.70.9: Note the vocabulary associated with human kingship: práśasti- ‘panegryic,
laud’ and balí- ‘tribute’. Since kings receive práśastis rather than bestowing them
and since the verb dhiṣe is medial (whatever else it is morphologically), the tr. of Old
SBE and Ge whereby Agni bestows the práśasti (e.g., Ge “du legst Wert auf die
Kühe, auf die Hölzer”) seem incorrect to me, esp. with the semantic bleaching of
práśasti to ‘worth, value’ (Old ‘excellence’). Although the phrase is distinctly odd, I
think the point of “you receive/acquire a laud among cows and firewood” is that both
the butter oblations represented by the cows and the firewood make noise when in
contact with fire, and this noise can be interpreted as an audible expression of praise.
For another instance of prá √śaṃs in Parāśara’s oeuvre, see I.73.2.
I.70.9–10: See Proferes (pp. 48–49) on these contrastive verses. Note the
complementary bháranta / bharanta beginning 9c and ending 10d respectively. Their
relationship and their semantic contrast is emphasized by the vi(…)s: bháranta
ví(śve)… ví … bharanta.
I.70.10: Neither of these hemistichs produces two proper Dvipadā Virāj pādas. The
first divides into 4 / 6 and presents as a Triṣṭubh lacking a syllable; the second has 11
syllables and is simply a Triṣṭubh, though with an irregular break.
I.70.11: What referent to supply with the adjectives sādhúḥ and gṛdhnúḥ isn’t
entirely clear. I follow Ge with ‘horse’, because sādhú- several times modifies horses
(see Ge’s cited parallels), though Re’s ‘warrior’ is equally plausible contextually.
As disc. in the art. cit., I do not believe, with Old (SBE, Noten) and others,
that this verse is a later addition. Rather it is a summary verse of the whole threehymn Vaiśvānara sequence, or even of all of Parāśara’s Dvipadā Virāj hymns to
Agni, and describes the Vaiśvānara as the ideal warrior and leader for the united
clans embarking on a joint enterprise.
I.71 Agni
113
I.71.1: The meter of this first verse almost serves as a transition from Parāśara’s
Dvipadā Virāj hymns that precede it to his Triṣṭubh ones beginning here (–I.73), as
all four pādas (as well as 2a) have openings of 5 and the first two are syntactic units.
The hymn then settles down into a pattern of mostly 4-syllable openings.
The underlying fem. subject of this verse is universally taken to be the fingers
of the priest producing fire by friction from the kindling sticks. The qualifier sánīḷāḥ
can be read with both the simile and the frame: the fingers belong to the same hand
as the co-wives do to the same household. So Ge, etc. The sexual innuendo fits the
friction context well, in addition to setting the stage for the incest theme to come.
The object phrase śyā́viṃ áruṣīm “the dusky and the ruddy [fem.]” is
generally taken to refer to night and dawn (although Old [SBE and Noten] wishes to
emend the text to fem. nom. pls. modifying the sisters, a generally bad idea). The
identification with night and dawn is certainly supported by III.55.11, adduced by Ge,
etc., with the same vocabulary. However, it would essentially duplicate the simile of
d (“they delighted in night and dawn, like cows in dawn”), a clumsiness that seems
uncharacteristic of Parāśara, and one also wonders why the fingers would care about
night and dawn. I think, instead, that this is another of Parāśara’s tricks, using color
terms associated with the two time periods to pick out another referent entirely,
namely the aráṇi- (fem.) ‘kindling sticks’, which the fingers should care about, since
they are manipulating them. The color differentiation is a little odd, but I don't think
it's entirely out of the question. In X.184.3 the kindling sticks are called 'golden'
(hiraṇyayī), and one can imagine that as the fire begins to catch hold, the upper stick
might glow red, while the lower one would be wreathed in dark smoke.
I.71.3: A difficult verse with a discouragingly large number of competing
interpretations. In outline, though not in detail, I follow Ge’s interpretation: pāda a
concerns the production of poetry and the poetic vision by our side, in contrast to the
poetic visions of the stranger, our rival, in b. After the stranger’s attempts are gotten
rid of, the topic returns in cd to our poetic products, which have success. (WG
basically follow Ge, while Old [SBE] and Re take b as coreferential with cd.)
The first hemistich, esp. its first pāda, is highly alliterative and rhyming:
dádhann ṛtáṃ dhanáyann asya dhītím, ā́d íd aryó didhiṣvò víbhṛtāḥ. The two
parallel 3rd pl. verbs dádhan and dhanáyan technically do not match in mood, since
dádhan should be a subjunctive (the injunctive to this redupl. pres. should be either
*dadhat or dadhur, and the subjunctive of the latter verb should be dhanáyān; see
Hoffmann 1967: 271 n. 13, who produces a tr. with subjunctive contrasting with
preterite). However, in this context I think dádhan was created as a nonce injunctive
to match the two an sequences in dhanáyan. The subjects of these verbs are
simultaneously the Aṅgirases and the current poet-ritualists; the injunctive
conveniently elides the difference between past and present action.
The beginning of b, ā́d íd, generally has temporal value (“just after that”) and
is so tr. in the publ. tr. But Parāśara uses the expression in causal value in I.67.8 and
I.68.3, and a causal value is possible in the next verse (71.4); such an interpretation
114
works better here, and I would therefore change the tr. to “because of that …” The
launching of their own poetic vision by our side scatters the visions of our opponents.
The sentiment, but not most of the vocabulary, echoes that in I.70.1, where the arí- is
also the opponent.
The desid. adj. didhiṣú- to √dhā is based on the middle value of that root
(‘acquire’); so Heenen p. 165, though I do not subscribe to the additional semantic
baggage Heenen attaches to the stem. It means ‘desirous to acquire’, hence ‘greedy,
voracious’. With most tr. I assume a pl. of dhītí- as the referent.
Just as the injunctives of pāda a can have either presential or preterital value,
the lack of verb in b allows the temporal value to be set by the preceding clause,
hence applicable both to the current ritual situation and its mythological model. The
publ. tr. might be changed to “are/were dispersed.” There is no agreement on what
víbhṛtra- means here, but ví √bhṛ generally means ‘disperse, pull apart, carry away’,
and the form also needs to be considered in conjunction with víbhṛtaḥ in the next vs.
The “unthirsting” dhītí- belonging to us/the Aṅgirases are implicitly
contrasted with the voracious ones of the arí-, as Ge, etc., point out.
With Gr, Old, Re, I take apásaḥ as a nom. pl. fem.; Ge (/WG) rather as a gen.
sg. masc. referring to the singer/priest. This is certainly possible and would provide a
neat contrast to the gen. sg. aryáḥ of b. I would prefer, however, to keep the possible
reference to the Aṅgirases alive.
I would prefer not to take devā́n in d as a gen. pl. construed with jánma (or as
a second acc. with vardháyantīḥ “strengthening the gods, their race”); therefore,
despite the pāda boundary, I take it with áchā, which frequently governs devā́n (so,
apparently, also Re); cf. esp. I.132.5 (=139.1) devām̐ áchā ná dhītáyaḥ, which is our
underlying phrase.
I.71.4: Another difficult verse, though it is clear that it alludes to the theft of fire by
Mātariśvan in the first hemistich. This theft is expressed through word play: the
verse begins máthīt, which can mean either ‘churn, rub’ (to produce fire) or ‘steal’.
In this Agni context, esp. given vs. 1, which concerns the churning of fire, ‘churn’
would be the most likely reading – and so it is tr. by Ge, Old (SBE), and Proferes
(2007: 31). But the name Mātariśvan ends the pāda, and this mythological allusion
would tip the balance towards ‘steal’ (so Re, WG).
The real problem comes in the middle of that pāda: the ppl. víbhṛtaḥ. This
form should mean ‘dispersed’ or ‘carried away’, and grammatically it should modify
mātaríśvā. However, sense would require it instead to qualify fire, the underlying
direct object in the clause; fire is often carried away and dispersed in various
locations. Cf. ví … bharanta in the preceding hymn I.70.10 and discussion there, as
well as the passages adduced by Old SBE, n. to vs. 4. There is no solution that
satisfies both sense and syntax. Ge suggests it's a transferred epithet, from Agni to
Mātariśvan; this seems a description of the problem, not a solution. Old suggests
simply displacing it rightward to read it with the assumed subj. of pāda b, namely
Agni (“When Mātariśvan had produced him by attrition, he … who was brought to
many places, has come to every house”), and WG apparently follow, but this kind of
115
extraction from one clause and insertion in another is not syntactically possible in my
opinion, esp. given that in its clause it is placed between the acc. referring to Agni
(īm) and the nom. mātaríśvā, so that its grammatical affiliation is emphasized. Narten
reports a clever suggestion of Hoffmann’s, that the ví- is really ‘bird’, and the form
means ‘carried by a bird’ (“Das vedische Verbum math, n. 38 [=KlSch p. 23 n. 38]),
but Mātariśvan should no more be carried by a bird than be dispersed. In the end the
least jarring rendering is Re’s “s’étant transporté çà et là,” presumably referring to
Mātariśvan’s journey. My “borne away” reflects a similar notion, and I also toyed
with the possibility that “carried away” may have the same extended meaning as the
English idiom, namely, “overcome with excitement.” No doubt Parāśara recognized
the trap he was setting, forcing the audience to find a way to interpret víbhṛtaḥ with
Mātariśvan rather than with the far more natural Agni.
As noted above, the subject of the next pāda must be Agni, established “in
every house” (cf. V.11.4, X.91.2). Because of the accent on bhū́t, the clause must be
parallel to pāda a and is not the main clause, which is introduced by ā́d īm in c.
On the second half of the verse see Proferes 2007: 30–31. Agni’s role as
messenger is of course well known: he travels between gods and men, heaven and
earth. This regular route may be prefigured here by Mātariśvan’s theft in a, for
elsewhere it is clear that M. stole fire from heaven (e.g., I.93.6). It is also said that M.
bestowed the stolen fire on Bhṛgu (see nearby I.60.1), and so the epithet bhṛ́gavāna-,
whatever its morphology, is apposite. What is unfamiliar is the simile, in which Agni
serves as messenger as if for a more powerful king. According to Proferes, Agni is
here “depicted … as the symbol of the presence of a leader’s authority within each
household.” I’m not sure, because Agni’s dūtá role involves traveling, not staying
put. I am inclined to think the simile was generated simply from dūtyàm and is not
meant to apply directly to Agni and his relationship to royal power.
I.71.5: This vivid account of the incestuous attack of Father Heaven on his own
daughter (elsewhere Dawn) comes as a surprise in this hymn. The connection must
be the fact that this incest led to the birth of the Aṅgirases, who were the subject of
vs. 2 (and possibly vs. 3); this birth is mentioned in the other incest verse, vs. 8cd,
though without naming the Aṅgirases. The other connection is that Agni, unnamed,
is the avenger in this verse, shooting Heaven as he relase his semen into his daughter.
Although the avenger is usually identified as Rudra on the basis of post-Ṛvic
passages (see Ge n. to vs. 5, Re, WG), I have demonstrated (Hyenas, 288–97) that the
original avenger was Agni. Note that ástā ‘archer’ is used of Agni in the last verse of
the immediately preceding hymn (I.71.11). I also take Agni as the subj. of pāda b:
like a hunter on his prey, he creeps up on the pair, having noticed the caresses
(pṛśanyàḥ) Heaven was bestowing on his daughter.
In addition I consider Agni to be the subject of pāda a (with most tr.), but the
action there is harder to explain: how and why does Agni make the sap=semen for
Heaven if he disapproves of Heaven’s sexual designs on his daughter and in fact
punishes him for the rape? For one thing, fire is unlikely to produce anything we
might consider sap; in a naturalistic sense, making rása for heaven would seem to be
116
storing up rain. However, our notions of nature and Vedic India’s are often at odds.
That the same semen (or possibly its delivery agent, the penis) is called tvíṣi- ‘spark,
flare’ in pāda d and, even more telling, is referred to as “blazing semen” (śúci rétaḥ)
in 8b, brings it into fire’s conceptual sphere. It may be that the flaring up of fire
produces Heaven’s semen. But what is Agni’s motivation? I’m afraid this remains
mysterious; perhaps it was involuntary, produced by the flaring up just suggested.
I.71.6: This verse returns us to the ritual here-and-now and is considerably easier to
interpret than the scraps of mythology in the previous two verses. Still, there are a
few puzzles.
The verb ví bhāti ‘radiates widely’ is surprising, since its subject must be the
mortal worshiper, even though Agni (or occasionally Dawn) is ordinarily the subject
of this common verb. Old (SBE, repeated in Noten) suggests emending it to ví dhāti
‘worships’ (in Old’s tr.). (I sometimes get the feeling that in the Noten Old felt the
need to stick with emendations he’d suggested in his far more free-wheeling SBE tr.,
even though in his later years he was otherwise very reluctant to emend.) But a
metaphorical sense of ‘radiate’ works fine in the passage, as Ge points out, citing
parallels like VI.5.5, also of a mortal worshiper. And this would be an instance like
víbhṛtaḥ in 4 in which the most natural referent of a word is not the grammatically
supported one.
Old (SBE and Noten) also proposes to read dat. uśaté in b, or rather suggests
that the underlying form in sandhi was a dat. uśata, since the spellings uśató ánu and
uśaté
ánu “belong to the inventions of Vedic grammarians” (SBE n. 2 to vs. 6). This is
certainly possible, but a gen. form uśató would reflect what we might call the
“proleptic” use of the genitive for dative to express indirect object: the reverence was
offered to you and so it is now yours. This colonization of the dative by the genitive
is very common both in Vedic and in Classical Sanskrit, and already here it may
reflect the weakening of the dative case that led to its disappearance in MIA. I
certainly see no reason to take the transmitted uśató as acc. pl. with dyū́n, as Old
(Noten) and Ge (n. to 6b) tentatively suggest.
I.71.7: The second hemistich consists of two contrasting halves: in c the speaker
complains that his (“our”) vitality (váyaḥ), the same vitality that Agni was said to
increase in 6c, is not to be seen throughout his kinsmen; this leaves Agni as the sole
figure who will find solicitude (prámati-) for him among the gods. Put thus, the
passage does not make much sense in English. It relies on the close relationship
between ‘solicitude’ and kinship (see esp. I.31.9-10, I.108.1, X.23.7, also adduced by
Ge) in the ṚV. The word prámati- denotes the kind of care and concern a father
shows for his children (or sons) and is regularly associated with, and indeed
identified with, the pitár- ‘father’ (see, e.g., I.31.9, 10, 14, 16) and secondarily with
the kin-group; cf. X.23.7 vidmā́ hí te prámatiṃ deva jāmivát “for we know your
(fatherly) solicitude, o god, like that of kin.” Since in our passage the jāmí- has
proved disappointing, the poet turns to Agni for satisfaction. The question then is
117
what is the relationship between vitality and solicitude, which seem conceptually
distant. This is less clear, but the connection is also found in the prámati-heavy I.31
already cited; cf. I.31.10ab tvám agne prámatis tvám pitā́si nas, tváṃ vayaskṛ́t táva
jāmáyo vayám “You, Agni, are (fatherly) solicitude, you a father to us. You are
creator of vitality; we are your kin.” In the immediately preceding verse (I.31.9)
Agni is also called tanūkṛ́t- ‘body-creator’. The balanced pairing of tanūkṛ́t- and
váyaskṛ́t- in a paternal context suggests that in producing children the father creates
both the physical body and the intangible essence of life that together make up a
living being – ‘life-force’ might be a reasonable rendering of váyas – and the father’s
prámati- seems to be the catalyst that brings them together. Moreover, at least on the
evidence of our passage, váyas- seems to be something that can be shared by the kingroup, but in this case is not, and in these circumstances if one cannot rely on the
shared vitality of his relatives, he must look only to his father and his father’s
solicitude. For another association of vitality and the father in Parāśara’s work, see
I.73.1.
The publ. tr. does not convey this very well. In c I would replace “is not
widely perceptible” with “is not widely evident” or “does not appear throughout.” I
also question my supplying an object (vitality) with cikitvā́n; I now think this simply
characterizes Agni as the perceptive one or indicates that he perceives the situation
as a whole. The same word cikitvā́n does seem to take an object in 5b, however.
Against the Pp. I prefer to take vidā́ as a lengthened 2nd sg. imperative rather
than subj. vidā́s, though there are no real implications either way.
I.71.8: The incest myth treated in vs. 5 returns here with no more clarity. The various
alternative treatments are too numerous to discuss here, so I will present only my
own version, some of whose details I have adopted from others. The first hemistich
describes again Agni as the avenger, loosing his weapon at Heaven in the midst of
the latter’s sexual encounter with his daughter. In pāda the téjaḥ, the sharp point of
the arrow, reaches the lord of men (=Agni, in my view), so that he can shoot it. I do
not understand why his preparation for shooting should be described thus, and in a
phrase ā́naṭ + ACC we might expect the acc. to express the target of the arrow.
Nonetheless, Heaven is not otherwise called nṛpáti-, and it would be odd to give him
this positive epithet in this situation. For Agni as an archer whose arrow has a téjassee VI.3.5 sá íd ásteva práti dhād asiṣyáñ, chíśīta téjò 'yasā ná dhā́rām “He, just like
an archer, has aimed (his arrow), about to shoot. He has whetted its point like a blade
of copper.” I take iṣé in this pāda as an infinitive to √iṣ ‘impel, send’ (others interpret
it as ‘to enjoy’); I then supply a form of this same verb in the next pāda, which
otherwise lacks one.
Note that the product of this incest, the young troop (śárdham … yúvānam), is
characterized as both ‘faultless’ (anavadyám) and ‘well-intentioned’ (svādhyàm),
asserting their goodness despite the circumstances of their birth.
I.71.9: As mentioned in the publ. intro., the reason for including this verse,
containing the Sun and Mitra and Varuṇa but no mention of Agni, isn’t clear.
118
However, it may be that Mitra and Varuṇa, the guardians of the moral and ethical
order, and their all-seeing eye, the Sun, who observes all behavior, are brought in as
witnesses of the shocking incest of the last verse and the appropriate punishment
inflicted by Agni.
I.71.10: Another apparent non sequitur. I have nothing to say about the morphology
of the famous crux vidúḥ, which occurs also in VII.18.1 in the same phrase with the
same close sandhi (abhí vidúṣ kavíḥ sán). I do consider it a nom. sg., not acc. pl.
I.72 Agni
I.72.1: The lexeme ní √kṛ, literally ‘make down’, idiomatically means to surpass
one’s opponent with regard to some quality by “putting” or “bringing” [them]
“down.” Indeed the English idiom “put down” is very close, but I chose not to use it
because it doesn’t convey quite the right sense with the right case frame. The verb
generally takes an acc. of the quality in question; the person who is outdone can be
either in the acc. (X.49.8), hence a double acc., or gen. (II.23.12), and there is some
disagreement among tr. about which construction we have here, since vedhásaḥ
śáśvataḥ can be either gen. sg. or acc. pl. Ge opts for the acc. pl. in the tr., but offers
the gen. sg. as alternative in his n.; Old, Re, WG choose the sg., as I have done,
though for no principled reason.
The náryā ‘manly powers’ in b contrasts both with kā́vyā in a and amṛ́tāni in
d. The first contrast would be between, roughly, warrior skills and verbal skills, the
second between human (though of course nṛ́- can often be used of gods, too) and
divine. For the latter pairing note that the two acc. pls. are construed with similar
verbs: middle participles to reduplicated formations, dádhānaḥ and cakrāṇáḥ.
I.72.2: After the general proclamation of Agni’s universal powers in vs. 1, the hymn
turns to the myth of the disappearance and recovery of Agni.
The immortals’ inability to find Agni “among us” (asmé) is contrasted with
the success of the mortal who finds him “on the highest track,” presumably in the
realm of the immortals, in vs. 4.
The acc. phrase vatsám pári ṣántam is ambiguous, since the participle can
either modify vatsám (“the calf enclosing …”) or take vatsám as object (“enclosing
the calf”). Ge (WG) interpret it as the latter, Old, Re the former. Ge is surely correct:
pári ṣántam elsewhere (IV.1.15, VI.17.5, adduced by Ge) qualifies the rock that is
the Vala cave and takes “cows” as its object; cf. VI.17.5c … ádrim pári gā́ḥ …
sántam “… the rock enclosing the cows.” Although in the usual myth of the
disappearance of Agni, he is hidden not in a rock but in the waters, this transference
of Vala phraseology is enabled here because Agni is called a calf, and cattle were
enclosed in the Vala cave.
There is slight phonological play in the phrase amṛ́tā ámūrāḥ. Also note víśve
amṛ́tāḥ “all the immortals,” a variant of víśve devā́ḥ. The use of this phrase is telling
because in Parāśara’s Vaiśvānara series (I.68.–70) víśve is reserved to evoke “all
119
(men).” See discussion esp. in I.68 and in Jamison forthcoming. I.68.8 also contains a
pāda-final ámūrāḥ, there referring to men.
Although the perfect of √sthā without preverb is almost always presential
(see Kü p. 580), in this mythological narrative it must be preterital (or at least so tr.
into English), like the injunctive vindan in pāda a; perhaps we can think of it as a sort
of “honorary” injunctive, whose temporal value is determined by context.
The form cā́ru (so Pp.; Saṃhitā cā́rv) is taken as a locative by Sāy (cāru
cāruṇi śobhane). This unlikely analysis is followed explicitly by Ge and Old (SBE,
more doubtfully Noten) and implicitly by Re and WG. (Gr. takes it as an adverb.) I
take it rather as the neut. acc. it appears to be and an alternative locational expression
to the adjacent loc. padé paramé, supplying nā́ma ‘name’ on the basis of nā́māni in
the next verse (3c) and the frequency of the collocation cā́ru nā́ma. It’s worth noting
that cā́ru nā́ma can be hidden or secret (e.g., II.35.11, IX.96.16), and so it would be
appropriate to this tale of the hidden Agni. What exactly it would mean to take their
stand at his name is not clear, perhaps at the place where Agni’s name is invoked in
ritual performance. This would fit with the following verse. It’s also possible that if
Agni’s dear name is what’s at issue, then the padé paramé might be the highest word,
as well as or instead of the highest track. (See 6ab below.)
I.72.3: For my interpr. of the purport of this verse in a Tānunāptra context, see publ.
intro.
The identity of the subjects in this verse is dispute: Ge “die Erzväter,” Sāy the
Maruts, Re the gods. I consider this vs. a continuation (of sorts) of vs. 2 with the
víśve amṛ́tāḥ still the subject. I have no explanation for the “three autumns.”
The interpretation of the verb form saparyā́n is quite problematic. It certainly
appears to be a subjunctive, and contrasts with the 3rd pl. injunctive saparyan that
Parāśara places in the same pāda-final position in I.70.10. A subjunctive does not sit
comfortably in the dependent clause whose main clause contains a preterital perfect
(dadhire) and an augmented imperfect (ásūdayanta). Ge (also Re) compares várdhān
in I.70.7 (which I take differently) and calls it the subjuctive “bei Zeitangabe” (n. to
I.70.7ab), whatever that means. He tr. it as a preterite. Hoffmann calls it a
“Konjunktiv in präterital Sachverhalt” (p. 244). Re “ont honoré et honoreront encore,”
attributing what seems to me a unique interpretation to the subjunctive. I consider the
form a straight preterital and have an admittedly quite ad hoc way to account of the
apparent subjunctive. The present participle to this stem, saparyánt-, is found 7x
pāda-final as nom. sg. masc. saparyán; I suggest that in order to make it clear that the
form here is not that pres. participle, the final vowel was lengthened, perhaps
redactionally. This lengthening has no metrical consequences, of course. The
lengthening did not happen in saparyan in I.70.10 because it was unaccented and
couldn’t be mistaken for a participle. I am quite unhappy with this explanation but
even more unhappy with the idea of a subjunctive in this context.
I.72.4: I consider ródasī a word play. It is of course a dual, referring to the two world
halves, but it also evokes rodasī́, the consort of the Maruts. Properly speaking she
120
should be in the acc. sg., hence a putative rodasī́m, but this form is not attested. In
I.167.4 we find rodasī́ in acc. usage: the singular consort has simply been
grammatically assimilated to the dual world halves in all her appearances.
The intensive part. vévidānāḥ could technically belong either to √vid ‘know’
or √vid ‘find’, whose middle generally means ‘acquire’, and it is somewhat difficult
to know which root is supposed to be represented in Ge’s “Gehör finden bei” and
Re’s “trouvant-accueil près de.” With Old and Schaeffer (p. 183–85) I take it with
‘find, acquire’, though I am somewhat puzzled by what the intensive part is supposed
to contribute – perhaps the sense of constantly keeping their acquisition close by; this
would work better for the consort than the two worlds. Schaeffer’s tr. (p. 185)
“nachdem sie die beiden hohen Rodasī gefunden haben” has no intensive nuance, and
she indeed suggests that the form does not really belong to the intentive, but is a
substitute for the perfect participle (*vividānā́ḥ) on metrical grounds, hence her
preterital tr. I find this substitution hypthesis unlikely, given that the intensive is
quite a marked formation and the accent patterns of the two formations are different.
The Pp. reads rudríyā (neut. pl.), not –āḥ with most tr. As Ge and Old (Noten)
point out, the Pp. reading is not impossible, but the nom. pl. works better. If the first
pāda makes reference to Rodasī, a nom. pl. referring to the Maruts makes better
sense.
Tr. differ on the interpretation of nemádhitā; I render it as close to my interpr.
of its other three occurrences, ‘when facing the other side’, used of battle arrays
(VI.33.4, VII.27.1, X.93.13). That it refers to men and gods being divided (so Ge,
WG) seems farfetched and ignores the evidence of the other identical occurrences
and inserts a referent (men and gods) that is simply not there.
The last pāda echoes 2d, except it is Agni standing on the highest track, not
his pursuers.
I.72.5: On this verse as an expression of the Tānunāptra ritual see Proferes 2007: 59.
The verse begins with a form of the resonant idiom sám √jñā ‘come to an
agreement, act in harmony” that was so important in Parāśara’s Vaiśvānara cycle
(see I.68.8, 69.9 and disc. there and in Jamison forthcoming). Here it expresses the
unity not of all men, as in the Vaiśvānara cycle, but of the gods, even including their
wives. Or so I interpret it; Ge and Re thinks the group includes both gods and men.
This seems unlikely to me, in part because pátnīvant- only qualifies gods (except for
pressed soma drinks in VIII.93.22). As I have discussed elsewhere, I do not think that
the Sacrificer’s Wife later called pátnī was yet established in the ṚV; she was only
being introduced (and this introduction disputed) in the latest strata of the text.
Most tr. take ririkvā́ṃsas tanvàḥ kṛṇvata svā́ḥ to mean “having abandoned
their former bodies, they took on their own (new bodies),” but this is somewhat
incoherent: were their old bodies any less “their own” than the putative new ones?
Proferes’s tr. (p. 59) seems self-contradictory, “Having yielded their bodies, they
made [their bodies] their own,” but his explanation makes sense of this: they
“overcome their atomization by relinquishing their ‘bodies’ in the course of a fire
rite; they thus make their own (svā́ḥ) the collectivity of their own individual bodies,
121
which is to say that each individual within the group identifies with the ‘bodies’ of all
the others.” By taking kṛṇvata as reciprocal, as I do, rather than just reflexive, this
sense can be found in the passage directly.
I.72.6: The Pp. reads unaccented avidan, but accented ávidan should be extracted
from the ambiguous sequence padā́vidan.
As often with RVic numerology, the identity of the three times seven secret
padā́ is not clear. Ge tr. padā́ as Wörte (sim. Re), which is certainly possible and has
parallels (see Ge’s disc. in n. to 6ab), but the padé paramé that figured significantly
earlier in the hymn (2d, 4d) should be kept in mind. Moreover, in the next vs. Agni is
said to be “inwardly knowing the roads (ádhvanaḥ) leading to the gods” (7c), and
since the padā́ were found in Agni here, it may be that these padā́ are tracks and
identical to the ádhvanaḥ that he inwardly knows. As I argued in the publ. intro., I
think the padā́ are the “tracks” of the ritual.
What amṛ́tam refers to is also unclear; it could be, as Sāy suggests, Agni, or
simply “immortality, the immortal principle” (so Re). It unfortunately cannot be the
body, which is fem. Note the “way towards immortality” (amṛtatvā́ya gātúm) in 9b
below.
The phrase sthātṝn
́ caráthaṃ ca with its number mismatch shows Parāśara’s
fondness for this merism (I.66.9, 68.1, 70.3, 7) and the grammatical anomaly found
in most of the occurrences in his oeuvre; see disc. ad locc.
I.72.7: The referent of “for them” in the tr. of b is the “settled peoples” of a.
On the possible relationship between the ádhvanaḥ here and the padā́ of 6b,
see disc. there.
I.72.8: The first pāda lacks an overt verb, but it is easy to read ví … ajānan from b
(with most tr.), rather than supplying an entirely different verb as Ge does
(“brachten”). The identify of the subject is likewise not given, but, with Ge, it is quite
likely the Aṅgirases. The qualifier svādhī́- ‘very attentive’ is a signature word of
Parāśara’s (I.67.2, 70.4, 71.8); the occurrence in 71.8 is applied to the troop,
presumably the Aṅgirases, that Agni begets in punishing Heaven’s incest.
The “seven youthful ones of heaven” are the heavenly rivers; note the
corresponding phrase in the previous hymn, I.71.7 (though in a simile) sravátaḥ
saptá yahvī́ḥ “the seven youthful streams,” and for divó yahvī́- III.1.6, 9, VII.70.3.
The position of ā́ between diváḥ and saptá yahvī́ḥ does give one pause, however,
since postposed ā́ with abl. can mean ‘from’. Perhaps this is a mixed construction:
“from heaven they discerned the … maidens of heaven.” I do not know what the
seven maidens of heaven are doing here.
The VP “discerned the doors of wealth” (rāyó dúro ví … ajānan) is a variant
of “opened the doors (of wealth)” found twice earlier in the Parāśara collection:
I.69.10 ví rāyá aurṇod dúraḥ … and I.70.10 duró vy ṛ̀ṇvan.
122
I.72.9: This verse is the most baffling in the hymn, both for its contents and for its
connection to the rest of the hymn. Ge has a very busy interpretation (followed by
WG) that involves the Aṅgirases and the Ādityas separating themselves from the
earth and contending to reach heaven. His interpr. rests on scraps of later mythology,
and it is very difficult for me to see where he finds contending parties in the passage,
much less the signs that would identify them as Aṅgirases and Ādityas. It also
requires an unlikely reading of ví √sthā as ‘separate’, rather than the usual ‘spread
out’ (as Re points out). All in all, the interpretation requires a superstructure that the
verse cannot support, and examining the passage without the presuppositions Ge
brings to it yields a very different picture. Unfortunately, however, the picture isn’t
appreciably clearer.
I think that the verse continues the theme of the original discovery of the
hidden “tracks” of ritual performance: having discovered them (see 6ab), the subjects
of 9ab, the gods most likely, perform the required ritual actions, which lead both to
offspring and to “immortality.” The second hemistich is considerably more difficult,
but I tentatively suggest that it also depicts a primal sacrifice. The “stretching” of the
ritual ground in an ordinary sacrifice, that is, the laying out of its boundaries and the
positioning of the offering fire, is here expressed in cosmic terms: the earth herself
spreads out to provide the ritual ground; she does this through the actions of her sons,
the ritual participants. And she then “suckles the bird.” Most take the bird as Agni,
whatever else they do with the verse, and the suckling, that is, the tending of the
ritual fire with wood and oblations, would work fine in this scenario. Re thinks rather
the sun, and again the sun as cosmic representative of the ritual fire is thinkable.
I will now treat some of the elements of the verse in more detail. Although
svapatyá- can be substantivized to mean “good descendants,” it is originally and
more commonly an adjective, and even when a head noun is absent, it can be
supplied. In this case I interpret the word in the same way as VII.91.3c víśvén náraḥ
svapatyā́ni cakruḥ “The men have done all (ritual actions) bringing good
descendants.” Here the governing verb is “mount” (ā́ … tasthuḥ), which I take as
metaphorical for ‘embark on’. With the journey theme of pāda b, the (actions of) the
sacrifice can be conceived of as a chariot, as so often in the RV.
The yé of ab is picked up by the “great sons” of cd (mahádbhiḥ … putraíḥ).
These can be the offspring produced by the ritual in pāda a, or more likely in my
opinion, the performers of those actions – the gods or, perhaps more narrowly, the
Ādityas – who are responsible for the cosmic sacrifice in which their mother, the
earth itself, spreads out as the ritual ground. I would slightly alter the publ. tr. to “by
means of her sons … the earth …” Keep in mind that the gods are the sons of Heaven
and Earth (cf. the bahuvrīhi in the dual devá-putre ‘having the gods as sons’), so
Earth is their mother. I am inclined against restricting the subject to the Ādityas. I
think rather that mother Earth is called Aditi because Aditi is an archetypal mother,
but her sons are all the gods, not just the Ādityas. The next verse (10b) simply refers
to the immortals, in what seems to be a continuation of this narrative, and the default
subject throughout seems to be the gods in general.
123
Note that “to suckle the bird” (dhā́yase véḥ) is a paradox that would be
recognized by any reasonably alert observer of nature (as the Vedic people certainly
were), in that birds aren’t mammals and don’t suckle. The sense of dhā́yase can of
course be bleached to something like Old’s (SBE) “for the refreshment of the bird,”
but I prefer to think this paradox was meant to be savored, along with the paradox of
the sons bringing about the action of their mother.
I.72.10: The first half of the verse may returns to the Tānunāptra theme. Although I
would prefer a middle verb to active ní dadhuḥ, the first pāda could depict the joint
deposit of their shared śrī́ in the ritual fire (cf. the echo of níhitā in 6b), while the
second pāda continues the cosmogonic theme of the previous verse. I confess,
however, that the very parallel I.73.4c ádhi dyumnáṃ ní dadhur bhū́ry asmin gives
me pause. Either I need to interpret that as also a Tānunāptra passage or delete the
reflexive implication here. Because of the active verb, I am inclined to change my
Tānunāptra interpretation and tr. simply “they deposited dear splendour in him,”
although since cā́ru- can have a quasi-reflexive sense ‘own dear’, it may be possible
to interpret the two passages differently.
As for the second hemistich, the rivers of pāda c are most likely the oblations
poured into the fire. The referent of the fem. pl. nī́cīḥ … áruṣīḥ is less clear, and in
fact most tr. (Ge, Old SBE, Re, WG) take the two feminines separately, with one
nom., the other acc. object of prá … ajānan. I think rather that the two words belong
together as subj. and refer again to the oblations. The lexeme prá √jñā does not
otherwise take an object, but just means “know the way, think ahead.” As underlying
referent of the feminine adjectives I would supply dhā́rā ‘stream’ vel sim.;
‘downward facing’ certainly applies to the oblations, though ‘ruddy’ is more difficult.
However, fem. áruṣī- can be used of cows; since the ghee poured into the fire is a
product of cows, it can be so described, even though ghee is of course not ruddy
itself.
I.73 Agni
I.73.1: For the complex relationship between ‘vitality’ (váyaḥ) and the father, see
disc. ad I.71.7. The “wealth acquired from one’s father” also reminds us of I.70.10,
where the sons divide and carry away the property of their aged father.
Agni’s traversing of the (sacrificial) seat (sádma … ví tārīt) in my opinion
refers to the removal of what is later called the Āhavanīya fire from the Gārhapatya
fire at the west end of the ritual ground and its ceremonial placement at the east end.
See 2d.
I.73.2: Pāda a is more literally “who, like Savitar, possesses realized thoughts,” but I
have adjusted the English to avoid awkwardness.
In c most tr. take satyáḥ as the shared term in the simile amátir ná (Ge “wahr
wie ein Bildnis”) or as modifying amátiḥ in the simile (WG “wie eine verwirklichte
Erscheinung”). The latter is impossible because amáti- is feminine. The former does
124
not convey much sense. The purport of the simile of course depends on the meaning
of amáti-, which I consider to be more concrete and specific than the usual glosses
‘appearance, form, picture’. An amáti- can be golden (III.38.8), broad and wide
(V.62.5, VII.38.2, 45.2); it is associated with lordship (V.69.1 amátiṃ kṣatríyasya),
can be displayed on a chariot box (I.64.9), and is unloosed or unfurled (V.65.2,
VII.45.3). All of this suggests that it is a pennant or ensign or other flag-like object.
On the basis of its association with lordship, I suggest that it can be a royal emblem
and that that is the basis for the simile here. As discussed ad I.70.9 and in Jamison
forthcoming, prá √śaṃs ‘proclaim, laud’ is a lexeme particularly appropriate to
kings, and here Agni is lauded by many like the royal emblem, which is the symbol
of the king. This interpretation leaves pāda-final satyáḥ somewhat stranded, and I
read it with the following pāda. Perhaps it was stationed in the c pāda because of
satyá(manmā) in almost the same position in the a pāda. Alternatively it can simply
be another qualifier of Agni in b, independent of the simile: “lauded by many like a
(royal) emblem, real …”
Most tr. give a rather vague and general rendering of the morphologically
elaborate desiderative gerundive didhiṣā́yyaḥ, “desirable to hold/win’. I think it has a
more technical ritual sense, referring to the installation of the fire (so also in the
other occurrence of this form in II.4.1). This would reflect the same sense of √dhā as
the adjective puróhita-, of the fire ‘placed/installed in front/to the east’ and the later
ritual complex the Agnyādheya ‘establishment of the fire(s)’. If my interpr. of 1d is
correct, namely that the offering fire has been taken from west to east, then the next
step would be its installation in the east.
I.73.3: For the first three pādas, cf. III.55.21.
The construction of this verse is more complex than the two preceding ones
and, in my opinion, displays some tricky relationships with Parāśara’s phraseology
elsewhere. To begin with, the first simile as it stands is pleonastic at best: Agni
should not be “like a god,” since he is a god. Nor should he be “like a god who
dwells on earth,” since in fact Agni is the only god, or (counting Soma) the primary
god, who dwells on earth. I suggest tentatively that the opening devó ná is there to
match ná rā́jā at the end of the hemistich. Even more tentatively I suggest that what
is really being compared is pṛthivī́ṃ viśvádhāyā(ḥ). Note first that in the preceding
hymn (10cd) mother earth spread out “to suckle the bird (probably Agni)” (dhā́yase
véḥ) and elsewhere earth is called “all-nourishing” (II.17.5 pṛthivī́ṃ viśvádhāyasam).
I think here that Parāśara is covertly comparing Agni to the earth with regard to this
well-known trait of hers, but does so playfully and allusively by stationing acc.
pṛthivī́m next to nom. viśvádhāyā(ḥ) (in the same metrical position as II.17.5 though
in Triṣṭubh not Jagatī cadence) and by the intertextual associations evoked by his
stating that earth suckled Agni in I.72.9. So the, or an, underlying meaning of pāda a
is “the god who, all-nourishing like the earth, …,” though both the position of the
simile marker and the case of ‘earth’ disallow this as a surface meaning. (Let me
make clear that I am not suggesting emending the text to *pṛthivī́.)
125
The two compounds hitá(mitra-) (b) and puraḥ(sádaḥ) may make implicit
reference to the word I suggested is to be associated with didhiṣā́yyaḥ in 2d, namely
puró-hita, as a descriptor of the installed ritual fire. The second, puraḥsád-, is
essentially a synonym of puróhita-.
Although I try to avoid explanations like “attraction,” I am afraid I must
follow Ge in taking the common term of the similes in c and d as “attracted” to the
comparandum: puráḥsadaḥ to the pl. vīrā́ḥ from putative sg. *puráḥsad, anavadyā́ to
the fem. nā́rī from putative masc. *anavadyáḥ. Although it would be possible to
avoid the attraction analysis by reading both terms as part of the simile proper (“like
heroes stationed in front and stationed for protection, like an irreproachable wife
pleasing to her husband,” so WG), this puts the simile marker one element too far to
the right and it also submerges the common term. This would be particularly
unfortunate in the first simile, since there is a play on two slightly different senses of
-sád- there. (In I.65.5 a similar gender mismatch is found, explained by Re as
attraction, an explanation I rejected. But there it does not involve a misplaced simile
marker.)
I.73.4: As noted at I.72.10 our c is almost identical to pāda a there. The publ. tr.
treats the first as a Tānunāptra passage (“deposited their own śrī́”) but not this one,
and the two should probably be brought into harmony. I now favor taking neither one
as a Tānunāptra expression, but see disc. ad I.72.10. In this case the deposit of
‘heavenly brilliance’ (dyumná-) may be what allows Agni to be come the foundation
of riches.
I.73.5: The verb ví … aśyuḥ should be read with both pādas, but, in my opinion, with
slightly different senses. In pāda a pṛ́kṣaḥ is a straight acc. object to the lexeme in the
meaning ‘reach, attain’; in pāda b víśvam ā́yuḥ may be one too (“attain a whole
lifetime”) as most take it, but it may also be an acc. of extent of time (“reach through
a whole lifetime”). The meaning is almost the same, but it would be like Parāśara to
put the constructions slightly off-balance, and the compound viśvā́yuḥ is something
of a signature word for Parāśara (see the immediately preceding vs. 4d, plus I.67.6,
10, 68.5) as an adverb expressing extent of time (“lifelong,” etc.). Note the phonetic
echoes, pāda a: ví … aśyur, b: ví … ā́yuḥ; the first of these distracted sequences also
evokes viśvā́yuḥ.
Pāda c sanéma … aryó is reminiscent of I.70.1a vanéma … aryó …
The phrase bhāgáṃ devéṣu … dádhānāḥ is almost a paraphrase of I.68.4a
bhájanta … devatvám. I would therefore change the publ. tr. “a portion among the
gods” to “a share in the gods.”
I.73.6: The referents of the “cows of truth” (ab) and the rivers (cd) are unclear, nor is
it clear whether they refer to the same thing. But I think they must be evaluated in the
context of this hymn, in which men (vs. 4), patrons (vs. 5), and gods (vs. 7) all serve
Agni in one way or another and seek rewards from him; the cows and the rivers then
must be another set of devotees and aspirants to his favor, esp. since the same phrase
126
sumatím bhíkṣamāṇā(ḥ) “seeking a portion of favor” is used of the rivers (6c) and the
worship-worthy (gods) (7a). I would tentatively suggest a twofold interpretation: 1)
A naturalistic one: the swelling cows are the rains, “apportioned by heaven,” with
their bellowing the thunder and the udders the clouds. The rains both are the
(heavenly) rivers and feed the (earthly) rivers. Recall the seven heavenly rivers of
I.71.7 and 72.8. Thus, the natural world pays heed to Agni, along with men and gods.
2) A ritual one: as Ge points out, these can be the milk and the water needed for the
soma sacrifice, though I confess I would prefer ritual substances more associated
with Agni. See I.72.10 where I suggested that the rivers were streams of ghee. I do
not believe that we need to see a reference to the Vala and Vṛtra myths here, as Ge
suggests.
I do not think that the rivers flow through the rock, pace Ge, etc., but rather
over. There are three passages in IX with samáyā and ví+VERB OF MOTION, all
dealing with soma going across the fleece (IX.75.4, 85.5, 97.56; cf. IX.85.5 vy
àvyáyaṃ samáyā vā́ram arṣasi “You rush across the sheep's fleece all at once.”). So I
think the rivers are flowing across or over the rock, but it would help tremendously if
I had any idea what the rock represents.
I.73.7: The voc. agne was inadvertently omitted from the publ. tr., so “o Agni”
should be inserted after “in you.”
The second pāda could also mean “they acquired fame in heaven.” The phrase
should be read with 5d bhāgáṃ devéṣu śrávase dádhānāḥ and 10c ádhi śrávo
devábhaktaṃ dádhānāḥ. The latter passage, especially, suggests that ‘acquire’ is the
better rendering.
The accented cakrúḥ in c can be explained, following Klein (DGRV I.176–77),
as triggered by the “rhetorically complementary nature of pādas c and d.” There is no
need to take the ca in c as a subordinator, particularly because subordinating ca
introduces conditional clauses. Note the contrast between ví(rūpe) and sáṃ (dhuḥ), a
common rhetorical pairing.
I.73.9: Note the echo of I.70.1, with optatives to the same (synchronic) root and a
likely identical case frame. See disc. ad I.70.1.
The second hemistich reprises material from earlier in the hymn: “wealth
acquired from their fathers” is exactly the same phrase, though in a different case, as
the simile that begins the hymn (I.73.1a); the patrons with their long lives were
encountered in 5b.
I.73.10: Ge’s interpr. of pāda c is persuasive. II.5.1 śakéma vājíno yámam (also
III.27.3, VIII.24.22) contains the missing horses supplied here. I take sudhúraḥ as
proleptic.
For pāda d cf. 5d and 7b; for devábhakta- cf. dyúbhakta- in 6b.
I.74 Agni
127
Several verses in this hymn are linked in relative cl. / main cl. structures: 1/2, 4/5. Vs.
6 may also be connected to 4/5. See disc. below.
I.74.2: This interpr. of snī́hitīṣu rests on a metaphorical use of the original sense of
√snih ‘snow’. On this root as warrior slang, see Hoffmann MSS 18 (1965) = Aufs.
447ff.
I.74.4: The second pāda contains two forms of √vī ‘pursue’, which by most tr. get
reduced to one, with the finite veḥ seemingly playing merely a modal or auxiliary
role: Ge “du … die Opferspenden entgegenzunehmen wünschest”; Re “tu vas pour
agréer ses oblations”; Old (SBE) “'to whose sacrificial food thou eagerly comest for
feasting.” The doubling is, however, captured by WG: “der du (die Götter) aufspürst,
damit sie die Opfergaben aufspüren,” resting on observations of Scar (498–99). I am
in independent agreement with WG on this construction. The root √vī regularly takes
both gods and oblation(s) as object; here the oblations are overt, but the gods are the
missing first object – the initial goal of Agni’s pursuit, to cause them in turn to
pursue the mortal worshiper’s oblations as underlying subject of the infinitive vītáye.
Cf. I.77.2 (also a Gotama hymn) agnír yád vér mártāya devā́n “When Agni, for the
sake of the mortal, has pursued the gods,” where the gods are surface object of véḥ
and the benefit for the mortal worshiper is emphasized. In our own hymn vs. 6 has a
full surface realization of the structure presumed here, with the gods as subj. of the
infinitive and the oblations its object, though with a different main verb: ā́ ca váhāsi
tā́m̐ ihá, devā́n … / havyā́ … vītáye “You will convey the gods here to pursue the
oblations.”
I.74.5: I recast the acc. to nom. in English, in order to be able to preserve verse
structure.
I.74.6: There is no obvious reason for the accent on the verb váhāsi. Ge (/WG), Re,
Klein (I.243-44) take it as subordinating ca, as does, somewhat tentatively, Old
(ZDMG 60: 733 = Kl Sch 208). But this verse does not work very well as a
conditional clause for vs. 7, and in fact Re recognizes this semantic disconnect by
ending his tr. of vs. 6 with suspension dots. If vs. 6 is subordinated to anything, it
would be better to connect it with vs. 5, repeating the message of vs. 4, which is
likewise subordinated to 5. I would simply call attention to several passages with this
same conjunction of elements with an accented verb and no clear motivation for the
accent: III.43.4 ā́ ca ... váhātaḥ, X.110.1 ā́ ca váha ... We may be dealing with a
catchphrase or with spread of the accent redactionally from a passage in which it was
correct to superficially similar phrases.
On the vītáye construction, see disc. ad vs. 4.
I.75 Agni
128
I.75.1–2: Note the play between the two final words of these verses: āsáni / sānasí.
The two verses also contain four superlatives in –tama, each to a stem ending in -as.
I.75.3–4: As noted in the publ. intro., only the first of the questions in vs. 3 is
answered in vs. 4, though the latter appears to be structured as the replies.
I.75.5: The three pādas contain three 2nd sg. verb forms belonging to √yaj, two pres.
imperatives yájā (a, b) and a so-called “si-imperative” yákṣi (c). Though “siimperatives” do often function as imperatives, they are haplologized -s-aorist
subjunctives by most accounts. I so tr. yákṣi here, since the poet could easily have
repeated yájā in this context; it would have been metrically fine and in fact would
have produced a more iambic pattern. It is accented because it follows an
extrasentential voc. For another si-imperative that is better rendered as a subjunctive
see ní … satsi in I.76.4 in the next hymn.
I.76 Agni
I.76–77: These two Triṣṭubh hymns are out of sequence, in that they both contain
five verses. The preceding hymn I.75 in Gāyatrī also contains five verses, and among
hymns of the same number of verses to the same divinity, those in the longer meter
(in this case Triṣṭubh) should precede. Curiously, Old does not remark on this. The
proper sequencing is restored with I.78, five verses in Gāyatrī.
I.76.1: As often, vā ‘or’ doesn’t present two balanced choices, but could be
paraphrased as “or, to put it another way…” That is, it rephrases and varies (often
substantially) a previous statement or question.
The first and last pādas contain forms of mánas-, which I have tr. differently.
The first, in the phrase mánaso várāya, is idiomatic in English as “heart’s desire,” not
“mind’s desire/choice/wish.” The expression in the last pāda, kéna … mánasā, might
better be rendered “in what spirit,” though I’ve chosen to stick closer to the ‘mind’
sense.
The pf. of √āp can be used presentially (so also Kü, though not with ref. to
this passage), and that value works best here: the focus in all four clauses is what we
can do to best serve Agni, not what someone has done in the past.
I.76.3: The accent on dhákṣi is unmotivated. WG label it as antithetical accent, but
that is only found in sequences that are more tightly bound rhetorically than this one.
(If all imperatives in sequence received “antithetical accent,” there would be many
more of them, beginning with ní ṣīda and bhavā in 2ab.)
Ge and Re both assert that the referent of asmai in d, the recipient of the guest
reception (ātithyám), is Agni. This seems perverse to me, though grammatically
possible. Agni is urged to bring Indra (“the lord of soma”) to the ritual ground; the
guest reception is surely for this new arrival and will consist of the usual ritual
129
offerings given to Indra, including soma. It is of course true that Agni is regularly
called ‘guest’ (átithi-), but that doesn’t seem relevant here.
The two hemistichs end with two unusual -van- forms: (abhiśasti-)pā́van- and
(su-)dāvan-. Though parallel in formation, they are in different cases (nom. and dat.
respectively), which somewhat conceals their morphological match. They are near
hapaxes: the first is found only once elsewhere in the RV (VII.11.3) in a similarly
constructed pāda, the latter only here.
I.76.4: Although the contents of this verse are pretty straightforward, the syntax is
particularly nasty.
The major problem comes in the first hemistich, with huvé, whose identity
and function are not clear. Or rather what it appears to be is not easily compatible
with the context. The form huvé and its unaccented counterpart are extremely
common and always 1st sg. mid. to √hū / hvā ‘call’, with the sense “I call / invoke X.”
If we take it in this sense and construe it with the nominative of pāda a, váhniḥ
‘conveyor’, there is semantic mismatch. The most likely referent of the expression
váhnir āsā́ “the conveyor by mouth” is Agni, who carries the oblations to the gods in
or with his mouth; see havyā́ júhvāna āsáni in the previous hymn (I.75.1) and the
next and final verse of this hymn where he performs sacrifice juhvā̀ “with his tongue.”
He should properly be the object of huvé. Old and Ge (/WG) interpret the váhnir āsā́
instead as the human poet, who brings the god(s) to the ritual by mouth, that is, by
his hymns. This is a clever idea and does allow the 1st ps. interpretation. The problem
then is the accent of huvé, though it could be classified with the problematic accent in
ā́ ca váhāsi in Gotama’s first hymn (I.74.6), which shares the opening ā́ ca. Re does
not want to give up the identification of the váhnir āsā́ with Agni (nor do I), and
attributes the syntax to anacoluthon: the first pāda is a nominative expression
describing Agni, which breaks off and gives way to an abrupt imperative “je
(l’)appelle.” The verbal accent remains a problem. I have another equally ad hoc
solution: that huvé is an infinitive (built like bhuvé to √bhū) and used predicatively,
as huvádhyai can be. This would solve the problem of accent, but the drawback is
inventing an otherwise unattested form that is homonymous with the extremely wellattested 1st sg. mid.
Note the sequence (vá)casā ...āsā́ (ā́) ca ... ca sa(tsī).
The other syntactic issue in the verse is the predicated vocative phrase in d:
bodhí prayantar janitar vásūnām. My tr. fails to render the predication; others (e.g.,
Ge) fail to render the voc. There is no good way to do this in English (or German).
Note that vásūnām is accented, although oblique cases that are part of voc. phrases
often are not (type sūno sahasaḥ “o son of strength”).
I.77 Agni
I.77.1: All tr. take the kṛṇóti in d as if it were ā́ kṛṇóti as in 2b, with the meaning
“attract the gods hither.” But the ā́ √kṛ idiom in vs. 2 is medial, and there is no ā́ here.
I prefer to take kṛṇóti as a dummy verb, standing for yájati, extracted from yájiṣṭhaḥ
130
‘best sacrificer’. Hence “does so,” that is “sacrifices.” This also helps account for the
íd.
I.77.2: The verse contains yet another example in Gotama’s collection of
anomalously accented verbs, here bódhati in the sequence sá cā bodhati. Cf. ā́ ca
váhāsi (I.74.6), ā́ ca huvé (I.76.4), as well as dhákṣi (I.76.3). Although it would be
desirable to have a uniform treatment of these verbs, esp. those following X ca, a
unitary account does not come to mind, and I have explained them in separate ways.
Here I suggest that we read the adverb sácā at the beginning of d and group it with c
(“… pursued the gods altogether”). Lubotsky gives only two examples of lengthened
cā, this one and III.57.5, out of over a thousand occurrences, so, as Old (Noten)
points out, the form is suspect. Old also considers but rejects a reading sácā, though
without giving reasons (beyond “kaum wahrscheinlich”). If sácā goes with the
preceding pāda, bódhāti begins a new clause and its accent is correct. Note that in the
previous hemistich (2ab), yáḥ …, hótā tám ū … ā́ kṛṇudhvam, the second pāda begins
with a nom. hótā that likewise belongs to the previous pāda, and a new clause begins
with the second word tám, whose clause-initial status is emphasized by following ū.
I.77.3: Note the verbal play in … ná bhūd ádbhut(asya) …
I.77.4: There are a number of interconnected difficulties in this verse, mostly focused
on what is happening in cd and the relation of this hemistich with the first one. I take
… ca yé maghávānaḥ … as the second part of an “X and which Y” construction with
naḥ in pāda a in Wackernagel’s Position: “… for us and (for those) who, our
benefactors …” Old (SBE), Re, and WG (if I am reading this last correctly) also
consider “us” and our benefactors to be the conjoined terms, though with differences
in detail. Ge, however, takes the benefactors as conjoined with Agni, and they should
also √vī our hymns and thought along with Agni. The frequency with which
maghávānaḥ (and also sūráyaḥ ‘patrons’) are conjoined with forms of the 1st pl.
favors the non-Ge interpr.
Another issue is the identity of the verb iṣáyanta in d. Ge seems to take it with
the fairly well-established stem iṣáyati ‘prospers, derives benefit’ (“… soll günstig
aufnehmen”), similarly Re (“jouissent”) and WG (“sich … erquicken”). With Old
(SBE) I take it as meaning ‘send, propel’ and ultimately a deverbal formation from
iṣnā́ti, via *iṣāyáti, of the type gṛbhṇā́ti, gṛbhāyáti, gṛbháyati.
The final problem is the form tánā. Ge (/WG) and Re take it as an instr. of
accompaniment “along with their offspring/descendents.” This is not impossible, but
this meaning is ordinarily found in the collocation tanvā́ tánā ca (e.g., VI.49.13). By
itself tánā occurs several times with girā́ (e.g., I.38.13, II.2.1) “with song at length,”
and I take it that way here as well, semantically construable with mánma ‘prayers’ at
the end of the hemistich.
I.78 Agni
131
See the publ. intro. for the structure of this rather dull hymn. Not only is the
third and final pāda identical in the first four (of five) verses, but in these same
verses the opening of the first pāda is also rigidly structured: abhí tvā (vs. 1), tám u
tvā (vss. 2-4). The remainder of the first pāda of these verses consists of two variant
pairs: gótamā girā́ (vs. 1), gótamo girā́ (vs. 2); vājasā́tamam (vs. 3), vṛtrahántamam.
It is thus only the second pāda of each verse that contains fully independent material.
The tvā in the various a pādas has a slightly different syntactic status in this
seemingly rigid schema. In vs. 1 it serves as the goal of the verb in pāda c, with
doubling of the preverb abhí (a, c). (Alternatively one can supply a verb in ab and an
object in c, as Ge (/WG) do, but this seems uneconomical. In the following two
verses it is governed by a verb in pāda b (2 duvasyati, 3 havāmahe), leaving nonumaḥ
in c without an expressed object. In vs. 4 it is again the goal of the verb in c, with an
apparent embedded relative clause in b.
I.78.4: The use of the epithet vṛtrahán-, esp. in the superlative, for Agni is of course
striking. The name agní- hasn’t yet figured in the hymn, but jā́tavedaḥ in vs. 1
clearly announces him as the recipient of this praise.
As noted above, the second pāda is technically an embedded relative
(assuming that nonumaḥ in c governs tvā in a).
I.78.5: As noted in the publ. intro., this final verse breaks the pattern of the rest of the
hymn, although it maintains the refrain. The aorist ávocāma “we have spoken” marks
this as a typical summary verse, referring to the activity in the rest of the hymn.
I.79 Agni
See the publ. intro. for the structure of this hymn (or rather the four hymns
collected here) and for an analysis of the first three verses.
I.79.1: As Ge notes, the identifications of Agni with sun (a), lightning (b), and fire
(cd) are Ludwig’s.
Most comm. take rájasaḥ as a subjective genitive with visāré (most clearly in
Ge’s “wenn der Raum sich weitet,” referring to morning in his view). I take it rather
to refer to the light of the sun’s rays (his golden hair) spreading through the midspace
at dawn.
The third pāda refers to the blazing up of the ritual fire at dawn. The fourth
pāda consists only of a simile, couched entirely in the feminine plural. The referent
of these feminine adjectives and the point of comparison with Agni are both unclear.
Ge supplies Frauen; Re suggests waters; Old (SBE) waters or dawns. The dawns
seem most likely (so also WG), since a sg. dawn already appears in c and they make
sense in this ritual context. Dawns are called apásas ‘busy’ in I.92.3, also a Gotama
hymn, and Uṣas is yáśasvatī in X.11.3. If the feminines are the dawns, the point of
comparison may be śúcibhrājāḥ ‘having bright blazing’, since the dawns are also
bright. If instead it is one of the adjectives in pāda d (‘glorious’ yáśasvatīḥ, ‘busy’
132
apasyúvaḥ, or ‘actually present’ satyā́ḥ), we must assume that an original masc. sg.
adj. has been attracted into the fem. pl. in the simile – not a problematic assumption).
I.79.2: Again, the interpretation of the image in pāda a depends on what the referent
is. Most (Ge, Re, WG) take suparṇā́(ḥ) as referring to Agni’s flames as birds. These
flames transform themselves (ā́ … aminanta) into storm clouds. As indicated in the
publ. intro., I instead think this verse develops the image of Agni as lightning found
in 1b. The suparṇā́ḥ are then the lightning flashes. This entails a somewhat bold
interpretation of the verb (though I would say no bolder than ‘transform themselves’,
for which no parallels are cited), from ‘change, exchange, alternate’ to the physical
image ‘zigzag’, from ‘change/alternate back and forth’. The middle of √mī is several
times used of night and dawn swapping their colors back and forth (I.96.5, 113.2),
which can be conceived of as a alternating pattern like zigzags. Against the flickering
light of the lightning there appears the black bull, that is, the thunder cloud, in b.
The nonce perfect nonāva, backformed to the intensive (see the repeated
nonumaḥ of the immed. preceding hymn I.78), I take as presential, with Narten (1981
“Vedisch lelā́ya,” p. 2 with nn. 7, 8, = KlSch. p. 234), against most comm., although
no harm would come of taking it as a preterite.
The pāda-final tag yádīmám is curious. It cannot be part of what precedes,
since nonāva is unaccented. The same phrase is found at IV.5.11, where I tr. “if it is
here.” I interpret it in the same fashion here, but in addition assume an enjambement
with the following pāda. Others keep it within its half verse (which I would prefer),
but in their interpretation the phrase seems like mere filler (e.g., Ge’s “wenn dies (so
ist)”).
The smiling females of c must be the lightning flashes (so also Ge, Re); cf.
V.52.6 vidyútaḥ … jájjhatīr iva “lightning flashes like giggling (girls).” Old (SBE):
rain showers, WG (tentatively): dawns. The “he,” subject of ā́gāt is the black bull of
b, the thundercloud as roaring fire.
For the semantic connection between these parts of the thunderstorm and the
aspects of fire see the publ. intro.
I.79.3: I take ab as a parallel and paraphrase of 2c, with Agni/the thunder cloud
leading (náyan) the lightning flashes, while himself swollen with rain. Others supply
different objects, and a radical (and in my view misguided) view of the passage
makes náyan a 3rd pl. finite verb with Agni as its object – quite awkward because he
is in the nominative in pāda a (píyānaḥ). The syntactic problem is resolved by
emendation (*píyānaṃ Old [SBE], subsequently rejected in Noten) or by taking
ṛtásya páyasā píyānaḥ as a parenthetical (Re).
It is unclear whether párijmā is meant as an epithet of Varuṇa or a separate
entity. Arguments can be made either way: in VII.40.6 the adjective qualifies wind
and is associated with rain, and having Wind (Vāta) involved in this thunderstorm
context would be understandable. On the other hand, a fourth god would break up the
unity of the standard Āditya triad. The same pāda is found in X.93.4, but in the
context of a longer list of gods.
133
Ge, Re take the “skin” as that of the earth, but I think it is rather the clouds in
the lower atmosphere. So, more or less, Old (SBE) and WG. Cf. nearby I.62.5
(attributed to a Gautama) divó rája úparam astabhāyaḥ “you propped up the nearer
realm of heaven.”
I.79.5: I take the point of the double utá construction (probably, with Klein DGRV
359, a nonce based on ca … ca) to be the contrast of the two time periods: at night
Agni must burn against the demons by himself (tmánā), but when dawn comes, her
light helps him out.
I.79.7: Since this is the first verse of a hymn in Gāyatrī (vss. 7-9), I take gāyatrásya
as a reference to a composition in that meter, rather than just a song.
On the basis of X.4.1 vándyo no háveṣu, I construe the loc. in c with the
gerundive.
I.79.9: I supply bhara from 8a; giving the matching verse openings, ā́ no, and the
same object ráyim, this seems the correct choice (so also Ge, Re, WG). It would also
be possible to make ráyim here the object of dhehi (so Old [SBE]). There are no
consequences either way.
I.79.10: The middle voice of the impv. bhárasva is correlated with the self-addressed
of the poet. See Jamison 2007: 104; 2009 [Skjaervo Fs.], esp. p. 70, and VII.88.1
I.79.11: The unusual post-verbal, pāda-final sáḥ is a nice match for the initial yáḥ
that it picks up.
I.80 Indra
I.80.1: Old takes śaśā(ḥ) to √śā ‘sharpen’ rather than √śās ‘order’ (as does Re,
judging from his portmanteau and barely comprehensible tr. “as-chassé-comme-enaiguisant”), but there seems no advantage in this. Neither root is found elsewhere
with níḥ, but a literal additive tr. of níḥ √śās works well here. And √śā does not have
a reduplicated form of this shape, whereas √śās has a pf. śaśāsa (etc.), to which this
can be a (plupf.) injunctive. So Kü (521), as well as Gr, etc. The níḥ found in vss. 2
and 4 (also 10), also of the explusion of Vṛtra, may account for the use of this
preverb here: it seems to be a signature word of this hymn.
I.80.3: As noted in the publ. intro., this verse represents a quoted example of the
“cheering on” alluded to in the refrain.
I.80.4: Ge suggests that the waters are “accompanied by the Maruts” (marútvatīḥ)
because the waters here are storm-rains. Although in a Marut context this would
seem reasonable, in this Vṛtra-smashing context the waters should not be rain but the
rivers confined by Vṛtra, and this interpretation is in fact possible here. Note that
134
Sarasvatī, the river par excellence, is called both marútvatī (II.30.8) and marútsakhā
(VII.96.2); this association must be a mediated one: the river is accompanied by or
the companion of the Maruts because their storm-rains swell the rivers.
I.80.5–6: I don’t see any real difference between áva jighnate (5) and ní jighnate (6)
that could account for difference in case frame: acc. in 5, loc. in 6.
I.80.7: In c I supply a preterital form of the redupl. jighnate of the last two verses
(*ájighnathās, though no forms with sec. endings exist to the stem), because the main
clause verb is the suppletive aorist to √han, avadhīḥ.
I.80.8: On pl. vájrāsaḥ see publ. intro.
I.80.9: With Re I take the numbers to refer to the people producing the songs, rather
than the songs or song-types themselves (as in the interpretation of Ge [/WG]). The
problem is that viṃśatíḥ is undeniably nominative, so the tr. “zu zwanzig” (and
parallel “zu tausend”) (Ge, [/WG]) misrepresents the grammar.
I.80.12: Ge tr. vépasā as “durch seinen Wortschwall,” with the verbal (“Wort-“)
component presumably on the basis of vípra- ‘inspired poet’. But near-adjacent
vépete in 11b, which refers only to physical trembling with no verbal component, is
surely the word against which to interpret it.
I.80.13: The expression in ab is striking, with Indra presiding over a fight between
Vṛtra and his (=Indra’s) missile, but such is the usage of yodháyati, as opposed to
yúdhyati ‘attacks’ (see Jamison 1983: 151)
I.80.14: The phrase yát sthā́ jágac ca is curiously formed, though the meaning is
clear. Because the verb rejate is unaccented, the yád cannot mark a “real”
subordinate clause. I wonder if it does not involve the imposition of an “X and which
Y” construction on an indivisible merism. In other words, with an “X and which Y”
we might expect *sthā yác ca jágat “the still and what is moving”; cf. X.88.4 (also
adduced by Ge) sthā́ jágad yád with a similarly pleonastic and apparently misplaced
yád. But sthā́ jágat is a fixed expression and nothing can be inserted into it.
Note the repetition of cit táva manyáve from 11a, again with a verb of
trembling (vépete 11b, vevijyáte 14d) and an instr. of fear (bhiyásā 11b, bhiyā́ 14d).
I.80.15: ná … káḥ must be read as a negative indefinite (“no one”) despite the
absence of an indefinitizing particle like cit.
I.80.16: Although the first hemistich begins with a rel. pron. (yā́m) and the second
with a form of sá/tám (tásmin), this is not a rel. cl. / main cl. structure, as the
difference in gender between yā́m and tásmin shows. The latter has Indra as referent,
and the former, which modifies dhíyam ‘poetic vision’ in b, is loosely picked up by
135
bráhmāni … ukthā́ “sacred formulations and hymns” in the main clause. In fact, I
think the referent slippage is deliberate, with bráhmāṇi being the culminating product
in our day of the age-old dhī́ of ab, attributed to legendary priest-poets. Recall that
this verse is in ring-compositional relationship with vs. 1, where a brahmán- creates
poetry in the new style.
I.81 Indra
I.81.2: see vs. 6 below.
I.81.3: Note the alliteration in b: dhṛṣṇáve dhīyate dhánā, a phrase that also contains
an example of neut. pl. with sg. verb (dhīyate dhánā).
I.81.6: This verse paraphrases and expands vs. 2. The nominal expression in 2b ási
bhū́ri parādadíḥ is turned into the verb parādádāti in 6b; śikṣasi in 2d is echoed by
śikṣatu in 6c; the phrase bhū́ri te vásu is repeated verbatim (2e, 6d); and the dative
beneficiaries in 2 (yájamānāya 2d, suvaté 2e) are replaced by dāśúṣe (6b) and, more
tellingly, asmábhyam (6c). The source of the wealth that Indra distributes is also
made clear: it belonged to the stranger and is presumably plunder.
I.81.7: Most tr. render ab in the 3rd ps., but there is no 3rd ps. verb here, and the
phrase is surrounded by 2nd ps. constructions (táva 6e, 2nd ps. impv. gṛbhāya 7c).
Nothing forbids an underlying tvám in ab.
Ge supplies “us” with śiśīhi, but, although this collocation does occur, I don’t
feel it’s necessary in this context.
I.82 Indra
I.82.1: artháyāse owes its accent to íd.
I.82.2: Most tr. take the two verbs ákṣann ámīmadanta as parallel (e.g., Ge “Sie
haben ja geschmaust, gezecht”), but the position of the hí after the second verb
strongly suggests that the second verb belongs to a separate clause that provides the
causal grounds for another clause. My translation reflects this: “for they have
brought themselves to exhilaration” explains what “they have eaten” actually refers
to: the consumption of soma or soma plus food-offerings. Then in b, as a
consequence of their having received nourishment from us, they repay us with goods.
Most tr. take priyā́(ḥ) in b as nom. pl. masc. referring to the Maruts; this
requires supplying an object for adhūṣata (e.g., Re “des biens”). I take it rather as an
acc. pl. fem., implicitly referring to a fem. noun like íṣaḥ ‘refreshments’; cf. X.134.3
áva tyā́ bṛhatī́r íṣaḥ … dhūnuhi “shake down these lofty refreshments,” with the
same VP.
With Old, etc., víprā must be a fem. instr. sg. modifying matī́.
136
I.82.3: The exact sense of the hapax pūrṇávandhura- ‘having a full chariot box’ is
not entirely clear. Ge suggests that it is full either because Indra has drunk so much
or because he is taking Opfergabe home with him; Re goes for the latter: “plein (de
biens).” This seems unlikely, since Indra is supposed to bring goods to distribute to
us, not take them away with him like party favors. Moreover the vandhúra- appears
to be the place where the charioteer stands (cf. the cmpd vandhureṣṭhā́-, as well as
I.139.4, III.14.3), not a container for cargo. Although a jocular interpretation like
Ge’s is possible, I wonder if a “full chariot box” is simply one that has the driver in
it; in other words, Indra has mounted the chariot and is ready to go. In any case,
pūrṇá- here has to be read in the context of the same word in the next verse.
I.82.4: I am puzzled by the reference here. The referent of cd, the one who makes
sure Indra’s cup is full, should be the sacrificer, but the sacrificer should not be
mounting Indra’s chariot. We might entertain the possibility that Indra is actually the
referent (and is looking out for his own interests by making sure the cup is full), but,
though switch between persons is very common in the RV, referring to the same
referent in both 2nd and 3rd persons in the same clause does not happen, as far as I
know – and there’s a voc. indra in the 3rd ps. yáḥ … cíketati clause.
I.82.6: With dadhiṣé we can supply either ‘reins’ or (as sometimes elsewhere with
√dhā and a loc. of ‘hand’) the vájra- (cf., e.g., VI.45.18 dhiṣvá vájraṃ gábhastyoḥ).
The presence of vajrin in d might support the latter interpr.
The Pp. reads amadaḥ in d. This causes interpretational difficulties: Indra was
ordered to leave the ritual ground and drive off to his wife in the immediately
preceding verse; he is still here in this verse, so how can he have already become
exhilarated in the company of his wife? Re must have recognized the problem, given
his tr. “tu t’es (toujours) enivré avec ta femme,” but this doesn’t work very well.
However, nothing prevents us from reading pátnyā madaḥ, with an injunctive, not an
augmented form. It is true that injunctive forms are quite rare to this stem, but this
does not seem to me a decisive argument. (I now see that WG also take it as an
injunctive, independently.)
I.83 Indra
I.83.1: Most tr. take vícetas- as ‘perceptive’ vel sim. (Ge “die verständigen (?)
Gewässer,” Re “les eaux discernantes”), but in this context it makes more sense to
use the intransitive sense of the root √cit ‘be perceptible, appear’ rather than the I/T
‘perceive’; hence, with ví, ‘widely perceptible, conspicuous’. The point of the simile
is surely that the goods with which Indra supplies the worshipper are abundant
enough to be easily visible, like the sparkling waters filling a river. Note the case
disharmony: logically the waters are compared not to Indra, but to the goods he
bestows.
137
I.83.2: This verse presents a number of small problems of interpretation, and it helps
first to determine what the verse is about in general. It seems to involve the gods’
ritual approach to the place of the soma pressing. They are compared to the goddess
waters in a because at the soma pressing waters are brought there to mix with the
pressed soma (cf. IX.69.4, e.g.). The identity of “the one who seeks the gods”
(devayú-) in c is not entirely clear, but my assumption is that it is soma. I also assume
that the acc. in d does not refer to the soma, but rather to the priest-poet, though it is
not impossible that the two accusatives have the same referent, most likely in that
case the priest-poet.
In b (with Ge [/WG]) I take the shared term of the simile to be vítatam
‘extended, extensive’. Both Ge and WG supply “Opfer” in the frame, but I would
suggest that it is, more specifically, the soma-filter, which is elsewhere so qualified.
Cf. IX.83.2 tápoṣ pavítraṃ vítataṃ divás padé, where the filter has cosmic
dimensions, as it would here.
In d brahmaprī́- is universally taken as ‘loving the formulation’, and this
interpr. would match that of brahma-dvíṣ- ‘hating the formulation’. However, both
here and in its other occurrence in I.152.6 I take it instead as ‘pleasing [someone]
with the formulation’ with the transitive value of act. prīṇā́ti, etc. In this passage it
makes more sense that the gods would delight in someone attempting to please them,
rather than someone who is himself deriving pleasure from something else. The
passage in I.152.6 invites a similar interpr.
I.83.3: With Ge (/WG) I take the two as the Hotar and Adhvaryu priests, with pāda a
appropriate to the former and b to the latter. Although it is something of a surprise to
find the charged word mithunā́, which is usually used of a sexual pairing, applied to
two males, it is presumably because the two priests have complementary duties. I do
not think that it refers to the sacrificer and his wife, contra Old (flg. Benfey). As I
have argued at length elsewhere (Jamison 2011, forthcoming a and b), I consider the
ritual Patnī to be a ritual role introduced only in the late RV and much disputed at
that time, and I doubt that she would be so casually alluded to here.
I.83.4: Assuming a thematic stem áṅgira- here is unavoidable, next to the far more
common s-stem áṅgiras-.
In order to give ā́d its usual temporal reading (“just after that”), I follow Old
in assuming that pāda b provides the grounds for the Aṅgirases’ acquisition of lifeforce.
Ge and Re supply a verb in d, but this seems unnecessary; the accusatives
here can simply expand on bhójanam in c.
I.83.5: This verse seems to jumble together a lot of primal mythology that does not
seem to be connected (or, rather, whose connection eludes me). As noted in the publ.
intro., vss. 4–5 describe the first institution of the sacrifice and touch on a number of
the primordial players. It is esp. noteworthy (but I don’t quite know what to do with
it) that vs. 4 contains the Aṅgirases and vs. 5 Atharvan, reminding us of the old
138
designation for the Atharva Veda, atharvāṅgirasaḥ (see, e.g., Bloomfield, Intro. to
Hymns of the Atharvarveda (SBE 42, 1897).
The form tate is a pseudo-perfect to √tan; at this period we would expect
*tatne (which we in fact get at X.130.2). It is metrically guaranteed. Kü (210) points
out that it is formed as if to a root √tā, and such a root morpheme could have been
extracted from the passive tāyáte. I would add that, as with many aberrant forms, the
context invites this form: note the immediately following word, opening the next
pāda, tátaḥ (recall also vítatam in 2b).
Note the chaining of ā́jani (b) … ājat (c) … jātám (d), with the middle term
belonging to a different root (√aj) from the first and last (√jan).
I.83.6: The relation between the three subordinate clauses (abc) and the main clause
is irregular in that tásya in d does not pick up a relative pronominal referent. I take it
as referring to the sacrificer who has been regularly present in the hymn (the presenttime vss. 1-3). Ge and Re may well be right that the kārú- ‘bard’ of c, which is
identified with the pressing stone, is its logical referent.
I.84 Indra
See the publ. intro. for the structure of this composite hymn.
I.84.1: Unlike the standard tr., WG take rájaḥ ‘Raum’ with the frame, not the simile:
“… soll dich (und) den Raum erfüllen, wie die Sonne …” I assume this is because
the simile particle ná, which usually follows the first word of a simile, here follows
the second word, by most interpretations (rájaḥ sū́ryo ná raśmíbhiḥ). This does not
seem to me sufficient reason to split apart this cosmological image. I attribute the
position of ná to the particularly close association of sū́rya- and raśmíbhiḥ, which are
frequently adjacent (cf. esp. the identical simile but without a third term … sū́ryo ná
raśmíbhiḥ in VIII.43.32, IX.41.5).
I.84.2: Note the chiasmic structure of cd: [ṛ́ṣīṇāṃ (GEN.) ca stutī́ḥ (ACC.)] úpa
[yajñáṃ (ACC.) ca mā́nuṣānām (GEN.)]. This could have been sketched in tr. by “up to
the seers’ praises and the sacrifice of the sons of Manu.”
I.84.7–9: Bloomfield (RR, ad I.7.8) suggests that each of these three verses reads like
a brahmodya, with the last two words (the four-syllable tag índro aṅgá, which he
considers to be prose “quite out of the metre”) serving as the answer. This analysis
words best for vs. 7, but his separation of the two-word tag from the rest of the verse
is clearly correct for all three verses -- though I do not think we need to consider
índro aṅgá “simple prose.”
I.84.7: I take índro aṅgá as the main clause corresponding to the rel. cl. introduced
by yáḥ.
139
I.84.8: The hapax kṣúmpa- is universally rendered by modern tr. as ‘mushroom’,
though there is no unanimity in earlier interpretations (cf., e.g., Gr: Staude [perennial
plant]). There is some etymological support for ‘mushroom’ from modern languages
(see EWA s.v.), and ‘mushroom’ works well in the simile, since kicking many
varieties of mushroom demolishes them, whereas a perennial plant is generally a
sturdier entity. I might go so far as to suggest specifically a puffball, since kicking
puffballs releases a satisfying cloud of dust (easily viewed on various YouTube
videos), and puffballs do appear to be found in the appropriate geographical area.
I.84.9: A broken construction. The subject of the 3rd ps. verb in c (patyate) must be
Indra, though he is represented by 2nd ps. tvā in the apparent rel. cl. of ab. (The
parallel in III.36.4b ugráṃ śávaḥ patyate dhṛṣṇv ójaḥ, where Indra is undeniably the
subject, makes it clear that the mortal soma-presser of our ab cannot be the subject of
patyate.) The yá- clause of ab thus has no direct grammatical connection with what
ought to be its main clause in c, and I therefore take yáś cid as the functional
equivalent of the indefinite káś cid.
I.84.10-12: See publ. intro. for the structure of this tṛca and its relation to I.80. The
subjects of all three verses are feminine (gauryàḥ, yā́ḥ … sayā́varīḥ 10; tā́(ḥ) …
pṛ́śnayaḥ, … dhenávaḥ 11; tā́(ḥ) 12; vásvīḥ 10-12), but, in my opinion, the hidden
referent of all three verses is the Maruts. See also comm. on vs. 16.
I.84.10: Ge (followed by Re, WG) takes the feminine plurals as referring to the milk
streams, mixing with the soma, here called honey. This of course accords well with
the feminine gender, and I agree that this is the first layer of reference. But both the
refrain, echoing the refrain of I.80, which has the Maruts as one set of subjects, and
sayā́varīḥ ‘fellow travellers, driving along with’ point to the Maruts as a second layer.
The adj. sayā́van- is used 3x (out of 6) with the instr. pl. of devá-, once (X.113.2)
quite clearly of the Maruts along with Indra. Why would the hyper-masculine Maruts
be referred to with feminines? I think the point is to conflate two of the sources of
Indra’s mastery: the soma-milk mixture animatized as cows, which enhances his
physical strength, and the Maruts, who provide verbal encouragement and moral
support.
I construe instr. vṛ́ṣṇā with sayā́varīḥ rather than with mádanti as most tr. do.
The latter is of course possible.
I.84.11: This verse contains another clue that the Maruts are the underlying referent.
The fem. subjects are identified as pṛ́śnayaḥ ‘dappled’, but pṛ́ṣṇi- is also the name of
the Maruts’ mother. They would be called Pṛśnis here, just as they are regularly
called Rudras, after their father Rudra.
Note the phonetic figure pr̥śanāyúvaḥ ... pṛ́śanayaḥ.
I.84.13: The use of the bones of Dadhyañc as a vajra-substitute is a particularly
puzzling part of this puzzling version of the Dadhyañc myth. The appearance of this
140
motif in the JB and MBh versions (see Ge n. on vss. 13–15) seems to me an afterthe-fact rationalization of the verse here.
I.84.14: The syntax of this verse is somewhat unclear. I take ab as a rel. clause with
the pres. part. ichán functioning as the main verb and an unusual, indeed disturbing,
position of the rel. pron. yád (we might expect *ichán yád áśvasya …), which is then
picked up by tád in the main clause of c. Most other tr. take ichán as part of the main
clause and the yád as the marker of an embedded rel. cl.: roughly “seeking the head
of the horse, which was set away in the mountains, he found it…” But not only
would I prefer not to allow embedded relatives in the RV, but the position of yád
makes this interpretation difficult, too (expect *ichánn áśvasya śíraḥ yád …?).
I.84.15: I confess myself to be entirely baffled by this verse, though the grammar is
straightforward. Ge’s reconstruction of the mythology (n. to vs. 15) is not entirely
compelling, nor is that found in WG.
I.84.16: The consensus of modern tr. is that the objects that the subject is struggling
to yoke are the priests (see esp. Ge), but I find this unlikely. The violent adjectival
descriptors seem uncharacteristic for priests, but quite suitable for the Maruts, to
whom the poet of this hymn (now drawing to a close) will dedicate the next four
hymns. Note esp. that both śímivant- and mayobhū́- are used of the Maruts, once in
the same hymn (VIII.20.3 and 24 respectively; for mayobhū́- see also I.166.3,
V.58.2). I therefore take the Maruts as referents of the acc. pls. and also suggest that
this verse is the pivot for the Marut reference found also (in my interpretation) in vss.
10–12. In that tṛca the Maruts are referred to in the guise of clearly feminine bovine
figures; here the cattle (gā́ḥ) are the first acc. object we encounter. Because gó- has
fluctuating gender, this form can of course be feminine (as the bovines were in 10–
12), but the immediate following adjectives establish it decisively as masculine. If
both 10–12 and this verse refer to the Maruts, this verse returns them to their proper
grammatical gender. Lüders (Varuṇa II.455) also thinks the Maruts are the referents.
ṛtásya can be construed with either dhurí or gā́ḥ, and tr. divide on which they
choose. I attach it to the chariot pole in part because ṛtásya is often construed with a
place (e.g., frequent ṛtásya yóni-), but the other is not impossible: cf. I.73.6 r̥tásya ...
dhenávaḥ.
I.84.17: Contra the interpretation of íbha- as ‘elephant’ in Vedic (so Ge), see EWA
s.v.
I.84.18: yajātai is, as far as I can tell, the only RVic examples of a medial 3rd sg.
subjunctive in -tai, the form that takes over beginning with the AV, spreading from
the 1st sg. As such it may be a sign that at least this part of this last hymn in the Indra
group is late.
Pāda c presents a problem that has been glossed over by most modern tr.: the
active of ā́ √vah should take an acc. of what is being conveyed, but it is distinctly
141
odd to say that the gods are bringing the oblation here. Ge (Re, WG) avoid the
difficulty by interpreting the verb as intrans. ‘fahren’ (/‘arriver-en-char’), a usage
associated with the middle, with hóma an acc. of goal. Old, however, disapproves of
this makeshift (as I do), setting out the arguments very clearly. Though he agrees that
“der Gedanke befremdet,” he sticks to the expected syntax (as do I) and cites a
number of passages in which gods bring oblations. The closest is V.41.7 in which
Night and Dawn bring the sacrifice to the mortal (V.41.7d ā́ hā vahato mártyāya
yajñám). I suggest that the role-switch in this verse (gods bring the oblation, rather
than coming to our oblation to take it away) is also found in the next verse, where the
god praises the mortal.
I.84.19: As in the last verse the usual ritual roles of god and mortal are reversed, with
Indra producing a praśasti of a mortal. I do not understand why, but, unlike 18c,
there is no way to wriggle out of the undeniable purport of this verb phrase – thus
supporting the “gods convey the oblation” interpretation of 18c.
tvám aṅgá reprises the tag of vss. 7–9, índro aṅgá, with of course the same
referent.
I follow Ge (/WG) in taking vs. 20 as the quoted vácaḥ of pāda d, though it
would be equally possible to interpret vácaḥ as referring to what precedes, indeed
even to the whole of the hymn.
I.84.20: It is tempting to take dabhan as related to or contextually assimilated to
dabhrá- ‘paltry, few’ in the meaning ‘come up short’, though the asmā́n would be
more difficult to construe.
I.85 Maruts
On the concatenative repetition in this hymn, see publ. intro.
I.85.1: Because of the frequent association of the root √śu(m)bh with verbs of motion,
esp. √yā ‘drive’ (cf., e.g., nearby I.88.2 śubhé káṃ yānti and cmpds śubhaṃ-yā́(van)), I interpret prá … śúmbhante more dynamically than most tr. (e.g., Ge “Die sich
wie Frauen schön machen…”), esp. given yā́man in the same clause.
The connection of the relative clause in pāda a with its presumed main clause
in d is interrupted by the hí clause in c. Ge convincingly suggests that the hí clause
gives an explanation or exemplification of the “wondrous power” attributed to them
in the last word of b, sudáṃsasaḥ, and he is followed by most tr. including me.
This clause contains a periphrastic causative cakriré vṛdhé with a medial
perfect as its base (see Zehnder, Das periphrastische Kausativ im Vedischen, pp. 23,
50–51). Although there is no doubt that the construction is a periphrasis, the reason
for its use is unclear. As Zehnder points out (pp. 23, 51), the imperfect of the wellattested morphological causative occurs with just this object in VIII.12.7 yát …
ródasī ávardhayat. The context here, which contains two present indicatives
(śúmbhante a, mádanti d), does not call for a perfect. I suspect (but cannot
demonstrate) that the use of this periphrasis with the perfect has something to do
142
with the middle voice of cakriré, which is also apparently unmotivated in the
periphrastic construction, and that there is an underlying pun -- on rodasī́, the consort
of the Maruts -- namely “the Maruts made the two worlds/*Rodasī their own” (ródasī
… cakriré), which would require a medial form. Since, as I demonstrated long ago
(“Voice Fluctuation in the Rig Veda: Medial 3rd Plural -anta in Active Paradigms”
IIJ 21 [1979]), 3rd pl. medial secondary forms in -anta to -áya-formations are almost
universally interpretable as re-marked actives in –an, using *avardhayanta here
would not allow the proposed double reading, with a true medial value in the pun. A
different pun is also possibly latent here, involving a reflexive reading of the middle:
“they made themselves grow strong” (cakriré vṛdhé). This interpretation would feed
naturally into the beginning of verse 2: tá ukṣitā́saḥ “once grown,” and 7a té
‘vardhanta svátavaso mahitvanā́ “Those self-powerful ones strengthened themselves
in their greatness” would echo this reflexive interpretation. In both of these
suggested puns, one of the words in the pāda has to be ignored (vṛdhé in the first,
ródasī in the second), but the suggestive if partial phraseology in both cases would
resonate with the audience.
It is also worth noting (though I don’t quite know what to do with this fact)
that of the fifteen occurrences of cakriré / cakrire in the RV, four of them are found
in this hymn (vss. 2, 7, 10 in addition to this one), a strikingly large percentage. And
that the three finite verbs in vs. 2 are all medial 3rd pl. with strong subject
involvement: pāda a mahimā́nam āśata “obtained (their own) greatness,” b cakrire
sádaḥ “made (their own) seat,” and c ádhi śríyo dadhire “put on (their own)
splendours.”
I.85.2: On the verbs in this verse, see final comments on vs. 1.
I.85.3: The pf. dadhire of b is not in harmony with the three pres. forms, śubháyante
a, bā́dhante b, rīyate c; in particular, the tenses of the subordinate-main clause dyad
in ab clash: “when they beautify themselves …, they have put on …” (The tr. avoids
the clash by using simply ‘put’, ambiguous between present and past in English, but
the Sanskrit is not so amenable.) It is difficult to wring a stative/resulative meaning
(“they have put on and now have on …”) out of dadhire, esp. given the identical
form in preterital use in 2d. It may be best to assume that dadhire here just echoes the
form in 2d, though it might be worth noting that the 3rd pl. mid. present indicative to
the redupl. pres., dádhate, which we might expect here, is quite rare. However, there
is another possible solution to this clash. The pres. śubháyante in the yád clause in
pāda a appears before vowel-initial añjíbhiḥ in the cadence, with, as is usual, the final
-e scanning short in hiatus (that is, probably as *-ay). It is possible that the underlying
form of the verb was actually the injunctive *śubháyanta and that the primary ending
-ante was substituted editorially to prevent the redactional contraction that might
have resulted (*śubháyantāñjíbhiḥ). This would allow a past-time reading of the first
hemistich, “When they beautified themselves …, they put on …”
143
I.85.4: This verse contains no main clause, just two different subordinate clauses
introduced by yé (ab) and yád (cd) respectively, with the first couched in the 3rd ps.,
the second in the 2nd ps. Although such switches of person in the middle of a verse
are far from rare in the RV (see in fact vs. 5), in this case it seems best to consider
the first half of this verse a pendant to vs. 3 (also in the 3rd ps.) and the second half an
anticipation of vs. 5 (also in the 2nd ps.), esp. since 5a is an abbreviated duplicate of
4cd.
The adj. manojúvaḥ ‘mind-swift’ could equally well be a nom. pl. masc.
modifying the Maruts or an acc. pl. fem. modifying pṛ́ṣatīḥ. Most tr. opt for the latter,
I think correctly on thematic grounds, but grammatically and semantically either is
possible. Gr and Macd (Vedic Reader) take it as a nominative; certainly its position
right before marutaḥ in pāda a, with pŕṣatīḥ appearing only in the middle of the
second pāda, suggests that the initial audience interpretation would be as a modifier
of the Maruts.
I.85.4–5: Verse 5 is an intrusive Triṣṭubh in this Jagatī hymn. (The final verse is also
in Triṣṭubh, but meter change is more usual in concluding verses.) The transition
between the meters is cleverly managed here (as Old already pointed out): the final
word of 4d (in Jagatī) and 5a (in Triṣṭubh) is áyugdhvam. The ending has the
distracted reading -dhuvam in 4d, but is simply -dhvam in 5a. The reason for the
change in meter is unclear to me.
I.85.5: As noted ad vs. 4, the person changes from 2nd to 3rd in the middle of this
verse. But this is not the only disruption: the main clause of 5cd begins with utá,
which has nothing to conjoin. Klein (DGRV I: 451) is puzzled and suggests,
somewhat desperately, that the utá “focuses on and emphasizes the second action of
the Maruts.” This seems to open the usage of utá unacceptably wide. It is possible
instead to see it as an example of “inverse utá,” conjoining the actions of the two
clauses of cd, with the parallel verbs ví ṣyanti and vy undanti (so also explicitly Macd,
VRS). Or it may be signaling the resumption of 3rd ps. discourse after the 2nd ps.
intrusions of 4cd–5ab.
The inundation of the skin in pāda 4 presumably refers to the wetting of hides
in the tanning process (so Ge).
I.85.6: In vs. 1 the Maruts were themselves called sáptayaḥ ‘spans’, but here that
image is “repaired” by separating the Maruts from the spans of draught animals that
bring them here.
The two raghu- cmpds. pick up raṃháyantaḥ in 5b.
The phrase urú vaḥ sádas kṛtám could technically be in apposition to barhíḥ
and hence accusative -- “Sit on the barhis, the broad seat made for you” -- but none
of the standard tr. so render it. The position of the vaḥ favors, but does not impose,
this nominal sentence interpretation. For another reason supporting a separation into
two clauses see comm. ad I.85.7.
144
The seat made for the Maruts on the ritual ground is here contrasted with the
one they made for themselves in heaven in 2b (diví … ádhi cakrire sádaḥ), which
event is then repeated in the next verse, 7b urú cakrire sádaḥ, though there the seat is
in/on/above the “vault” (nā́kam).
I.85.7: I tr. avardhanta as reflexive, rather than (with most tr.) intransitive ‘grew
strong’, in part because svátavas- suggests that their power comes by their own
efforts and in part because of its resonance with one interpretation of 1c, for which
see above.
On pāda b see comments ad 6c and the similar phrase in 2b. The accent of the
apparent main verb tasthúḥ is surprising. Macd (Reader) invokes the principal that
the first of two antithetical verbs is accented, but this seems a feeble explanation of
this particular situation. Ordinarily such antithetical verbs are adjacent to each other
and the semantic antithesis is clearer: “they mounted” and “they made” do not seem
particularly antithetical. I see two possible explanations: either the ā́ … tasthúḥ
clause should be taken as an unmarked subordinate clause (“[when] they mounted the
vault, they made …”) or the “antithetical” explanation is correct, but the verb to
which it is antithetical is not following crkrire. Instead it is found in 6c, whose two
clauses are in patterned contrast to 7b:
6c sī́datā́ barhír
urú vah sádas kṛtám
7c ā́ nā́kaṃ tasthúr
urú cakrire sádaḥ
The second part of both pādas contains urú sádaḥ √kṛ; the first parts contain the
preverb ā́, main verbs built to the semantically oppositional roots √sad ‘sit’ (sī́data)
and √sthā ‘stand’ (tasthúḥ), and an acc. of goal, again oppositional, in that the barhíḥ
of 6c is on the earthly ritual ground and the nā́kam is in heaven. I therefore think it
likely that the accent on tasthúḥ is meant to signal the contrastive relationship
between the two pādas.
The Pp. analyses dhā́vad as ha ā́vat, and most tr. follow this interpretation, as
containing an imperfect of √av ‘help’ (Ge “Als Viṣṇu dem … Bullen … beistand…”;
Re “eut aidé”; Macd “helped”). (An injunctive ávat would also be possible.)
However, we are likely dealing with a double, or even a triple, word play here, for
dhā́vat can also simply be the 3rd sg. injunctive belonging to the root (or rather roots)
√dhāv ‘rinse’ and ‘run’. This is recognized by WG, who translate with the second,
“Als Viṣṇu zu dem vom Rausch bewegten Stier läuft…,” but in the notes also raise
the possibility of ‘rinse’, referring to Gotō (Präsensklasse p. 186 n. 333), who in turn
refers to Pirart (IIJ 27 [1984] 102ff.).
I think that the two primary readings are “rinsed” and “aided” and that the acc.
object vṛ́ṣaṇaṃ madacyútam has different referents depending on the interpretation
of the verb, namely soma and Indra respectively. This ambiguity is made possible by
the fact that vṛ́ṣan- ‘bull’ is very commonly used of both Indra and soma. The
modifying cmpd. madacyút- also has two different interpretations, enabled by the
functional ambiguity of root noun compounds like this. Root noun second members
regularly have transitive force, governing their first members, hence here “arousing
exhilaration.” And in fact this compound is common in this sense, modifying soma
145
(IX.12.3, etc.). But passive value of the root noun with an instrumental or other
oblique sense of the 1st member is also possible, hence here “roused by the
exhilarating drink / roused to exhilaration.” For general disc. see Scarlatta pp. 128–
29. In this second sense Indra would be the obvious referent, as he is elsewhere (e.g.,
I.51.2).
While ‘rinse’ may strike the casual reader as an odd choice of words, in fact
rinsing is one of the standard steps in the preparation of soma. For Viṣṇu’s
participation in this process, see VI.17.11cd pūṣā́ víṣṇus trī́ṇi sárāṃsi dhāvan,
vr̥trahánam madíram aṃśúm asmai “Pūṣan, Viṣṇu (and the others) rinse the
exhilarating Vr̥tra-smashing plant, three lakes (full), for him [=Indra].” In the ‘aided’
interpretation, Indra is the referent of the accusative and the allusion is to Viṣṇu’s
help given to him in various exploits including the Vṛtra-smashing. In the “ran to”
reading, which strikes me as the least interesting, presumably Indra is also the
referent of the accusative, though WG don’t make this exactly clear.
I.85.9: The array of tenses in this verse requires comment. The subordinate-main
clause dyad of ab/c contains an imperfect ávartayat in the yád clause and a present
dhatté (dhattá in sandhi) in the main clause. Pāda d, which describes actions that
necessarily follow the one in pāda c, then has two imperfects (áhan and aubjat). I
consider the imperfect / present combination in the first sentence to be a makeshift
attempt to convey anteriority in a language without a semantic pluperfect. That is,
there is no structural means in Rigvedic Sanskrit to convey past anterior action
(“when he had X-ed, he Y-ed”), since the “pluperfect” is simply the past tense to
presential perfects. In later Sanskrit the gerund serves as a non-finite way to express
the value (“having X-ed, he Y-ed”), but the gerund barely exists in the RV. Here the
relative tense values are conveyed by the imperfect followed by the present, since the
imperfect does express action prior to that of the present tense. The imperfects in
pāda d then re-establish the past-time context. I therefore tr. the imperfect and
present of abc with the English pluperfect and preterite.
On náry ápāṃsi see disc. ad VIII.96.19.
I.85.11: Old suggests a clever, but ultimately unsatisfactory, reading in pāda a: rather
than táyā diśā́ with the Pp., he divides tá(y) ādíśā, with the instr. of ādíś- and the
older sandhi form of té ‘they’. This would make the end of the pāda more parallel to
10a … tá ójasā. He thinks that the Maruts performed the action in 10a with brute
force, but that in 11a “durch klugen Anschlag.” Although I am drawn to this idea
because of the persistent parallelisms in this hymn, I cannot accept it in the end. For
one thing the one hand ādíś- does not occur in the instr., and it generally means ‘aim,
intention’, which does not fit here. Old’s proposed reading also involves altering the
accent from diśā́ to ādíśā, so that it is not merely a matter of redividing the Saṃhitā
text. There is also the serious question of why tái̯ would be preserved in one
prevocalic environment but not in another, in adjacent vss.
I.86 Maruts
146
On the structure of this hymn and its syntactic patterns, see publ. intro.
I.86.1: I do not understand the function of hí here, which is doubly unusual in
occurring both in a relative clause and in the initial clause of a hymn. Hettrich
(Hypotaxe, p. 181) also finds it puzzling.
The abl. diváḥ is construed by Ge and Re with the voc. vimahasaḥ (Ge “ihr
ausgezeichneten (Mannen) des Himmels”), but we might expect it to lose its accent
in the vocative phrase, like divo naraḥ (II.36.2, V.54.10) and divo duhitar (IV.51.10,
etc.). It is better taken with √pā ‘protect’, which is elsewhere found with the ablative
(e.g., X.158.1 sū́ryo no divás pātu), indicating the location of the protector and hence
the direction from which the protection comes. So also WG.
I.86.2–3: As discussed in the publ. intro., the syntax of these two verses is to be
interpreted in the template provided by vs. 1, namely a relative clause (or clauses)
expressing the beneficiary of the Maruts’ favor with the genitive yásya and a main
clause assigning a reward to him. The full structure is seen in 1ab (rel. cl. with yásya)
/ 1c main clause with coreferential sá. Verses 2 and 3ab serve as the relative clauses
to the main clause of 3c, also beginning sá. However, the relative pronoun is
suppressed until 3a and the structure of parallel relative clauses is only conveyed by
the repeated vā ‘or’ (2a, 2b, 3a). Indeed, though 2b contains a genitive of the human
beneficiary (víprasya), 2a lacks even that: we must infer a yásya to limit the yajñaíḥ,
as well as a verb to construe with that instrumental. I would diagram the structure as
follows, with what is to be supplied in parentheses:
template, vs. 1:
yásya … pāthā́ (2nd ps. pl. verb) (1ab)
for whom … you Maruts perform benevolent action (1c)
/ sá
/ he is (rewarded)
complex variation, vss. 2-3
(yásya) yajñaíḥ vā (2nd ps. verb) (2a)
or by (whose) sacrifices (you were attracted?)
(yásya) víprasya vā … śṛṇutā́ (2bc)
or of (which) poet … you heard
utá vā yásya vājínaḥ … átakṣata (3ab)
or of which prize-seeker … you fashioned
/ sá
/ he is (rewarded)
This interpretation differs significantly from that of other tr. In particular, vs. 2 is
generally taken as an independent sentence that does not parallel the yásya structures
of vss. 1 and 3 and that has śṛṇutā́ as the verb of a main clause, not of a subordinate
clause, as I take it. (By that account śṛṇutā́ is accented because it follows the extraclausal vocative that opens the pāda.) Although my interpretation requires several
elements to be supplied, esp. in vs. 2, most other tr. also supply a verb with 2ab, and
147
they fail to capture the structural parallelism that allows the ellipses to be filled in a
principled fashion.
I.86.2: My interpretation requires matīnā́m to be construed with hávam “the call of
the thoughts”; for this expression see, e.g., VI.69.4 hávanā matīnā́m.
I.86.3: As noted in the publ. intro., I take the vājín- here to be the patron, for whom
the Maruts create a worthy poet.
I take gántā as a periphrastic future. VII.32.10 gámat sá gómati vrajé is
entirely parallel, save for having a subjunctive (gámat) in place of the agent noun
here, which suggests a future sense for the latter.
I.86.4–5: As indicated in the publ. intro., the genitive of the human worshiper is
continued in these verses, though with the demonstrative, not relative pronoun.
I.86.5: The condensed expression of the earlier part of the hymn continues here, and
there is no consensus on how to construe pāda c, which has no overtly signaled
connection with the rest of the verse. For example, Ge takes it as a simile, marked by
cid, with its comparandum in pāda b (roughly, “who dominates the peoples like the
flowing nourishments the sun”). But even if cid could mark similes (and I don’t think
it can), the simile doesn’t make sense. I will not rehash here the various possibilities
floated by other interpretations. My own generally follows Renou’s in supplying
asyá from 1a as the oblique predicate of a possessive nominal sentence: “(his) are …”
It would also be possible to take the perfect participle sasrúṣīḥ as the predicate: so
WG “Sogar zur Sonne sind (seine) Nahrungen gelaufen (erreicht),” but the exact
value of such a predicated perfect participle eludes me.
I.86.6: There is an abrupt shift from the unidentified 3rd sg. worshiper rewarded for
his work who dominated vss. 1–5 to the 1st ps. plural, but the rhetorical structure
invites the audience to identify the us of vs. 6 with the 3rd sg. of the earlier verse. The
dadāśimá ‘we have done pious work’, combined with the causal hí ‘for’, seems to
offer a summary of the activities of the previous verses. The repetition of the pl.
carṣaní- in 5b and 6c also supports this identification: he who “dominated” the
carṣaní- in 5b can be the same as the “we” who accomplished what we have done
with the help of the carṣaní- in 6c. This lexical chaining should have been signaled in
the translation, where instead the two words are rendered differently: “bordered
domains” (5b), “separate peoples” (6c). I would now use “separate peoples” also for
5b.
I.86.7: The 3rd sg. protagonist and the syntactic structure of vss. 1–5 return here, but
with the syntax reversed: sá … yásya.
I.86.8: I take the vā here as inverse, connecting the two double gen. phrases
śaśamānásya … svédasya and kā́masya vénataḥ. Klein (DGRV II: 205) also thinks an
148
inverse reading is possible (though he doesn’t use the term), but he also suggests vā
here might be equivalent to vaí. This seems unnecessary, and the inverse
interpretation gets some support from the inversion of the syntax in vs. 7.
Note the chiastic structure of the two genitive phrases: in the first the personal
participle śaśamānásya depends on the material svédasya, while the likewise
personal participle vénataḥ depends on kā́masya.
I.86.9: The first two pādas show a nice syntactic conversion: the 2nd member śávasof the voc. bahuvrīhi satya-śavas- ‘having real strength’ (unaccented, but would be
satyá-śavas-) is implicitly extracted from the compound and represented by tát,
object of āvíṣ karta.
The third pāda has a striking phonetic figure vídhyatā vidyútā, which was
already anticipated by pāda-initial vidā́ in 8c.
I.86.10: Another phonetic figure in a: gū́hatā guhyàm, whose -atā also echoes
vídhyatā.
In the publ. tr. guhyàm would have been better rendered as “to be concealed”
than “concealable,” since the concealment is not merely possible but desirable.
I.87 Maruts
I.87.1: Ge tr. usrā́(ḥ) as “Sternbilder (?)” because he considers the interpretation with
dawns as “kein naturwahres Bild.” But at early dawn stars are still visible.
I.87.2: On upahvará- see disc. ad I.62.6.
I.87.4: As noted in the publ. intro., in contradistinction to all standard tr. I take the
gaṇáḥ of d to refer to the poet Gautama, whose patronymic is Rāhūgaṇa and consider
the 2nd sg. reference of the 2nd hemistich as Gautama’s self-address. The other tr.
assume that the flock of Maruts is being addressed.
All standard tr. also take pāda d as a single clause (e.g., Ge “darum ist die
bullenhafte Schar ein Gönner dieser Dichtung”). But this interpretation fails to
account for the medial position of áthā, which is otherwise almost invariably initial.
(Klein, DGRV II: 70–71, notes the anomalous position here but follows the standard
tr.) I take áthā as clause initial, proclaiming Gotama’s identification with the Marut
flock on the basis of the shared qualities stated previously.
I.87.5: See the publ. intro. for discussion of the contents of this verse.
It may not be too farfetched to note, in connection with pāda d where the
Maruts acquire “their names worthy of worship” (nā́māni yajñíyāni), that in 1d they
were explicitly unidentified and undifferentiated (ké cid “whoever they are”). The
acquisition of these names follows their “reaching” (ā́śata) Indra, presumably to
render him moral support in the Vṛtra battle with their chants (ṛ́k-) (pāda c). In other
treatments of the aftermath of this battle they successfully negotiate with Indra for a
149
share in the soma sacrifice (see I.165, etc.), and the expression “acquired names
worthy of worship / sacrificial names” may be a way of expressing this incorporation
of them into the dedicands of the soma sacrifice. Note that in VI.48.21 the Maruts
assume Indra’s sacrificial name (dadhire nā́ma yajñíyam), also in connection with
the Vṛtra battle.
I.87.6: It is odd that the Maruts, just called ṛ́kvan- themselves (5c), here join
themselves with ṛ́kvabhiḥ. Ge suggests either that they have an “Anhängerschaft” of
singers or have gone among singers themselves; Re that it’s an instrumental of
identification (“en tant que chantres”). Gr invents, for this passage alone, a meaning
‘Glanz, Strahl’ for ṛ́kvan-. Given the identification of the human poet with the
Maruts in the same capacity in vss. 4–5, the singers of 6b may be human singers,
although the other instrumental attributes in this verse do not fit this picture. Ge
points out that in the first verse of the next hymn, I.88.1b, the Maruts’ chariots are
svarká- ‘equipped with lovely chants’, and so the ṛ́kva- here might refer to their
chariots, which are conspicuously absent from the list in pādas a–c. This is the
explanation I currently prefer.
The word iṣmín- is glossed by Gr as ‘eilend, stürmend’ (followed by WG),
while Ge and Re take it as ‘possessing arrows’, deriving apparently from a
suggestion of Bloomfield (see EWA s.v.), a suggestion emphatically rejected by
Mayrhofer (loc. cit.). Certainly a derivation from íṣu- ‘arrow’ is, to say the least, not
without problems. But in two of its other three occurrences, iṣmín- is found in a
weapons context as it is here, and the derivation therefore seems worth attempting.
Starting from íṣu-, the -in- possessive would be *iṣvín-. It is possible that the normal
distribution of -vant- and -mant- possessives, with -mant- regularly built to stems in u- (paśu-mánt-, etc.), as opposed to -vant-, would make *iṣvín- beside íṣu- appear
anomalous, and it was “corrected” to iṣmín-. Note the regularly formed íṣumant- (2x).
I.88 Maruts
I.88.3 As indicated in the publ. tr., I take the so-far-unnamed poets (who appear as
the Gotamas in the next verse) as the subject of kṛṇavante. They are erecting their
thoughts like trees for the Maruts to chop down with the axes in pāda a. By contrast,
most tr. take the Maruts as subject of kṛṇavante (so explicitly Re, implicitly WG; Ge
hesitates between the Maruts and the poets).
The simile in pāda b causes some minor grammatical problems. The Pp reads
medhā́, but if this word belongs to medhā́- (fem.) ‘wise thought’, we should rather
read medhā́ḥ (so Ge, Gr.). This is of course possible in this sandhi context, but pādafinal ūrdhvā́ is not amenable to such an analysis: it must be neuter plural and
therefore can’t modify a fem. medhā́ḥ. Gr sets up a neut. medhá- ‘Schaft’ for just this
passage, presumably to deal with the apparent gender disharmony. But it is simpler
to assume that ūrdha- has been attracted into the neuter by the vánā in the simile (so
Old).
150
The epithet tuvidyumná- is otherwise only used of gods, save for IX.98.1,
where it modifies the wealth brought by Soma. The only other plural form (V.87.7)
modifies the Maruts, but the standard tr. assume that it here qualifies the priests,
because the Maruts appear in the dative in pāda c and should therefore not be the
subjects of the verb in the same clause. The difficulty disappears if we read c with b,
rather than d. I then take the Maruts to be the subject in d, with the epithet
appropriate to them, and also take the middle dhanayante as intransitive with an acc.
of goal, rather than transitive “set the (pressing) stone in motion” with other tr.
I.88.4: For the imagery here, see the publ. intro.
The sequence páry ā́ va ā́gur with doubled preverb is curious. Ge provides
numerous parallels in his n. 4a, but none like this, with the two preverbs separated
only by an enclitic pronoun and contained with a preverb (pári) – verb sequence. I
have no explanation for this, though I would note that of the 8 occurrences of
supposed aguḥ / águḥ, 7 are pāda-final and are univerbated with the preverb ā́, as
ā́guḥ.
The hapax vārkāriyā́m is of course very obscure. Ge simply takes it as an
unanalyzable name of the “Sangeskunst” of the Gotamas. However, it is clearly a
compound and the compound members are easily identified: vār- ‘water’ and √kṛ
‘make’, and so it should be susceptible to meaningful analysis. As indicated in the
publ. intro. I believe that it has the same underlying referent as anubhartrī́ in 6a,
namely the musical instrument, the vīṇā, and that the poet was playing with the
paradox that the instrument is grammatically feminine but in some forms looks
remarkably like male genitalia. I therefore follow Gr’s gloss ‘Wasser … schaffend’
and assume that it refers to the penis. The problem is what kind of formation it is,
and how to get an agentive-like meaning out of kāriyā́-. First, note that the formation
does not have gerundive accent and is therefore not a gerundive despite the vṛddhi in
the root and the shape of the suffix. Formally it is most likely a verbal abstract
‘water-making, Wasserbeschaffung’, as Old takes it (so also AiG II.2.832 following
Old). However, because it was a feminine abstract, I think that it was secondarily
available to be identified with the underlying feminine referent, the vīṇā, and
animatized as a quasi-agent.
Note the phonological play between vārkār- (b) and arkaír (c).
I.88.5: All standard tr. take ná as negating the verb aceti. It seems unlikely to me that
an annunciatory aorist would be negated (“this very thing has not just been seen”),
and its clause-internal position, not immediately preceding the verb, also seems
unusual for such an interpretation. I take it instead as negating only the following
word yójanam and expressing the surprise that what has just appeared is very close,
not (even) a trek away.
I thus take yójana- as a measure of distance, as it generally is, rather than as
“Gespann,” with most tr. Rather than yójanam I take the referent of etát tyát to be the
formulation (bráhma) that the Gotamas made in the previous verse. This formulation,
also referred to as a dhī́- ‘insight’ in 4b, providentially appears at the same moment
151
as the Maruts approach. In what sense does the formulation “appear”? I interpret this
verb to mean that the poem that Gotama created in private (sasvár 5b) is now being
performed in public (i.e., at the ritual welcoming the Maruts). This interpretation also
entails supplying the verb “made” in 5b (echoing kṛṇvantaḥ in 4c), rather than “saw”
(with most tr., anticipating páśyan 5c).
My last departure from the standard tr. in this verse is in taking the acc. pl.
masculines in cd, híraṇyacakrān, áyodaṃṣṭrān … varā́hūn as comprising two
separate but conjoined NPs without overt conjunction (as so often), rather than one
single NP, since I find “golden-wheeled boars” an unlikely entity even in the RVic
universe of discourse. With ‘golden-wheeled’ in c it is natural to supply ‘chariots’;
the “copper-tusked boars” in d can easily be an extravagant characterization of the
Maruts.
I.88.6: For detailed disc. of this verse, see my 1981 article, “A Vedic sexual pun:
ástobhayat, anubhartrī, and RV I.88.6,” Acta Orientalia 42 (1981[82]) 55-63. The
gist of the article is the interpretation of anubhartrī́, a feminine agent noun built to
ánu √bhṛ, a euphemistic idiom that refers to sexual penetration -- e.g., in the cosmic
incest myth (X.61.5). The paradox of creating a feminine agent noun from this idiom
is the trick of the verse and echoes the use of vārkāryā́- in vs. 4 -- both, in my view,
referring to the vīṇā, a feminine noun but a musical instrument with a
characteristically phallic shape. The hapax causative ástobhayat ‘cause to sound’
(beside the simplex stobhati in b) has as its implicit subject the (male) player of the
instrument and taps into the (probably universal) notion of a man playing a woman
like an instrument in sexual encounters, with the added fun of the gender reversal in
anubhartrī́. The instrument both sounds (práti ṣṭobhati, b) and is caused to sound
(ástobhayat, c), a reasonable dual view of what a musical instrument does. I take
vṛ́thā āsām, with its unidentified fem. gen. pl., as referring to all women, who, in
male fantasy anyway, all “want it.” -- conveyed by my somewhat loose tr. “the way
women like it.” Other tr. supply ‘songs’ or ‘chants’ with āsām, for which there is no
particular support.
I.89 All Gods
I.89.1: The subjunctive of √as plus infinitival vṛdhé, lit. “will be for the
increasing/strengthening of…,” i.e., “will be (ready/available/amenable) to
increase/strengthen” is found also in 5c: 1c vṛdhé ásan, 5c ásad vṛdhé.
I.89.2: A form of devá- is stationed at the beginning of every pāda, the first three
being the gen. pl. devā́nām -- an effect difficult to capture in English without
awkwardness.
I.89.4: Ge (/WG) take the repeated tád’s of a-c (but not the one in d) as referring to
the mayobhú … bheṣajám “the remedy that is joy itself” and supply verbs in pāda bc
to support this object. Re by contrast takes the repeated tád as adverbial, ‘ainsi’. This
152
is more or less what I arrived at, though ascribing somewhat more meaning to the
adverbially used neuter pronoun. I assume it refers to the call embodied in the nivídof 3a (though nivíd- itself is fem.), which call is finally the suppressed object of tád
… śṛṇutam “listen to this” in d.
I.89.6: The rigid parallel structure of the four pādas is resolved by the final word of
the verse dadhātu, which verb must be supplied for the first three pādas. The rigid
structure, svastí naḥ GOD EPITHET, also allows the metrical irregularity of the first
two pādas to be kept under strict control. Under HvN’s interpretation each of those
two pādas has a rest after the suvastí naḥ opening, with the GOD EPITHET phrase taking
the rest of the line. Old suggests reading trisyllabic indira, as often, as well as pū̂ṣā́,
with distracted ū. The former seems more likely than the latter, but because of the
parallelism of a and b in HvN’s reading, I prefer that one.
I.89.7: The identify of the mánavaḥ in c is not clear. I take it as a continuation of the
description of the Maruts, though it is the case that mánu- and its derivatives do not
characterize the Maruts. Ge (/WG) suggests that it refers to men who have become or
been assimilated to the gods, like the Ṛbhus, while Re thinks c belongs with d and
refers to the All Gods, though this explanation runs into the same problem as the
Marut identification.
I.89.10: Pāda b is a textbook example of gender attraction of the pronoun in
equational clauses. Since the referent of the pronoun is feminine Aditi, we might
expect *sā́ pitā́ *sā́ putraḥ, but the gender of the predicated noun is transferred to the
pronoun.
I.90 All Gods
I.90.2: On the small class of -āna-nominals to -u-stems (with guṇa of the suffixal
vowel), mostly personal names, see AiG II.2.275.
I.90.4: It is unclear from Ge’s ‘aussuchen’ (/WG’s ‘auswählen’) whether they
attribute ví … cyantu to √ci ‘pile’ or √ci ‘observe’, since the proposed meaning
would require metaphorical stretching for either root. With Gr (and, judging from his
‘dégager’, also Re), I take it to the ‘pile’ root, with the literal sense being ‘pile
[obstructions, detritus, etc.] apart or away’, thus ‘clear’, of paths. This idiom is found
several times elsewhere of paths: IV.37.7, IV.55.4, VI.53.4 (passages collected also
by Ge ad IV.55.4).
I.90.6: Supplying ‘blow’ as the verb of a not only conforms to universal good sense,
but is suppored by vs. 4a of the previous hymn (I.89) vā́taḥ ... vā́tu.
I.91 Soma
153
I.91.1: The multiple possible meanings of the root(s) √cit and the unclear
morphological identity of (prá) cikitaḥ make interpretation of the first pāda
somewhat difficult. I follow Thieme’s solution (Plusq.), adopted also by Kü (Pf. 17677), that it is underlyingly a medial injunctive to the perfect stem (that is, an
unaugmented pluperfect), patterning with pf. cikité, etc. But the underspecification of
this posited *cikita caused it to be re-marked with an active ending (like root presents
of the type aduha-t). It should not be a subjunctive, despite its thematic appearance,
because of the zero-grade root syllable, and, pace WG, it should not be a trans./caus.
reduplicated aorist because of the light reduplicating syllable (expect *cīkitas) -although I do have to admit that acikitat in VII.80.2 does seem to function like a
redupl. aorist. WG supply panthā́m, from b, as object in a, but this seems
unnecessary. On a potential ring made by prá cikitaḥ here and prá cikitsā in the last
pāda of the hymn (23d), see publ. intro.
The other question in pāda a is whose inspired thought is at issue. I assume
that it is ours (that is, the poets’), in that the priests and poets create the ritual that
makes soma manifest.
I.91.2: This verse is structured by a series of etymological figures involving a nom.
sg. modifying soma and an instr. pl. specifying his qualities -- a: krátubhiḥ sukrátuḥ,
b: dákṣaiḥ sudákṣaḥ, c: vṛ́ṣā vṛṣatvébhiḥ, d: dyumnébhir dyumnī́. It is a not a subtle
device, but effective. On the first hemistich see further ad vs. 14.
I.91.3: Soma is here identified with the three principal Ādityas.
The (pseudo-)gerundive dakṣā́yya- has attracted a number of different
renderings -- in this passage alone, Ge “zu Gunst geneigt,” Re “qu’on doit-servirefficacement,” WG “als Zufriedenzustellender.” I prefer to tie it more closely with
dákṣa- ‘skill’, as “whose skill is to be sought,” despite the awkwardness of the
English gloss. In this passage Soma was just credited with skill (2b) and will later
dispense it (7c; cf. also 14c), and in both I.129.2 and VII.97.8 the ability of Indra and
Bṛhaspati to bring about good things is emphasized. JPB in VII.1.2 prefers “to be
skillfully tended,” quite close to Re’s rendering above. This latter tr. would work for
Agni (II.4.3, VII.1.2), who requires ritual tending, but less well for Indra, Bṛhaspati,
and (here) Aryaman.
I.91.4: The perennially difficult dhā́man-, found also in 3b, is a bit difficult to
interpret here as well. In both 3b and 4ab, the dhā́man- appear to be concrete and
locatable in space, for which the tr. ‘domain(s)’ fits well (cf. also WG “Stätten”). But
then in cd Soma is urged to accept our oblations with them, which seems difficult to
do with a place and edges closer to Ge’s “Formen,” a nebulous and all-purpose
rendering that I try to avoid with this word. Re’s “structures” doesn’t help either. I
must assume that “accept with all your (domains)” is a compressed way of saying
“wherever you are, accept.” The same problem is found, to some extent, in vs. 19.
I.91.6: A good example of subordinating ca.
154
I.91.7: As Ge and Re both point out, the dat. yū́ne ‘youth’ in b suggests that the
parallel mahé in a refers specifically to an adult.
I.91.10: I take the pf. part. jujuṣāṇáḥ as expressing an action anterior to that of the
main verb, impv. upā́gahi. If this is correct, the verse is constructed chiastically, with
imáṃ yajñám construed with upā́gahi, while the call that precedes Soma’s arrival at
the sacrifice is nested in between (idáṃ váco, jujuṣānáḥ).
I.91.11–12: The two alliterative root noun compounds vacovíd- (11b) and vasuvíd(12b) nonetheless contain the two synchronically separate roots, ‘know’ and ‘find’.
I.91.14: Although it is tempting to take dákṣa- as adjectival here (so Ge, WG), I am
somewhat dubious that this stem can be an adjective, and in any case the emphasis on
the skill associated with Soma in this hymn (vss. 2b, 3d, 7c) suggests a nominal
reading here. Although the pairing of an abstract quality (skill) with an animate being
(poet) might seem awkward, I see it as a variant reprise of 2ab, where krátu(‘intention’) and dákṣa- were paired; as discussed ad I.2.7–9, krátu- and dákṣa- are
two of the three qualities required to bring an action about and are regularly
associated. Here kaví- stands in for krátu-. This substitution is enabled by the
syntagm “the poet’s krátu-, as in the cmpd. kaví-kratu-. (Re’s interpr. is similar; see
his n.)
I.91.19: Here again, as in vs. 4, the dhā́māni of Soma are recipients of the oblation;
Ge’s Formen again works more smoothly, but I still consider these to be the locales
listed in 4ab, each of which can be a site of sacrifice.
Note the disharmony in number between the neut. pl. subj. (tā́ viśvā) and the
sg. pred. nom. (paribhū́ḥ) and verb (astu). This could be an example of the inherited
construction of neuter pl. with a singular verb, but I think it’s more likely just a
constructio ad sensum, with “all these” summarizing the plural subjects of the rel. cl.
as if in a mass. English “all this” can be similarly used, after a listing of discrete
entities. The Sanskrit has just been more punctilious about maintaining number
agreement.
Ge, Re, WG all supply ‘lifetime’ as the implicit object of pratáraṇaḥ, rather
than my ‘us’; they are most likely correct, given how common the idiom ā́yus- prá
√tṝ is. I might change the publ. tr. to “lengthening (our lifetime).”
I.91.20: Ge (/WG) takes the final phrase of d yó dádāśad asmai as expressing the
indirect object with dadāti in b (“gives [to him,] who …”), rather than as a
qualification of vīrám (or even of pitṛ-, as Re almost seems to). I think this is correct,
but it is mildly disturbing that there is no overt expression of the datival recipient in
the main clause; we would expect tásmai (see Ge’s n. 20d). It may have been gapped
because of the asmai in the rel. clause. Note that asmai cannot be a part of the main
155
clause and refer to the indirect object, the man who does pious service, because it is
unaccented; it must be part of the rel. cl. and refer to Soma.
I.91.21: This verse plays a minor syntactic trick. It consists of a string of acc. sg.
masc. qualifiers; since the previous vs. also consists largely of an acc. sg. masc. NP
headed by vīrám, the audience would be likely to assume that the new string of
grammatically matching adjectives are also qualifiers of vīrám, esp. since this type of
syntactic dependence between verses is not uncommon in the RV. It is only when we
arrive at the middle of the final pāda that we encounter tvā́m and discover that the
accusatives of this verse refer to Soma, not to the hero he gives us in 22.
Ge accounts for the curious hapax bhareṣu-jā́- ‘born at raids’ by the fact that
soma is especially offered before battle. This is probably correct, though the
semantics could be tighter. Scar essentially accepts this explanation and cites III.51.8
(of soma) jātám … mahé bhárāya, though he explores some other possibilities as
well.
I.91.23: In the phrase devéna … mánasā we again confront a noun (devá-) that seems
to be used adjectivally (see vs. 14 above). I have half given in to this temptation, with
“god(like).”
The verb ā́ tanat in c poses some difficulties in interpretation. Surprisingly,
none of the standard tr. who discuss the phrase notes that the same lexeme (ā́ √tan)
appears in the previous verse (22c) in the same metrical position and in a common,
almost cliché́d usage. I think that ā́ tanat here is a slangy expression, deliberately
constrasting with the high-style cosmic-description usage of ā́ tatantha in 22. I take
the subject to be the wealth (or share of wealth) of b, rather than the mánas- of a, as
Ge (/WG) take it, nor would I following Re in taking it as an impersonal construction.
It is not entirely clear who the “both” are in d: Ge singers and patrons, Re
men and gods.
Ge supplies ‘path’ with prá cikitsā (“sei … der Pfadfinder”) on the basis of
IV.47.20. This is possible, and might even be supported by the panthā́m in vs. 1,
since prá cikitsā seems to form a slight ring with prá cikitaḥ in 1a. However, the verb
does not need an object, in my opinion; the desiderative here can express a general
intention to be alertly perceptive.
I.92 Dawn
As indicated in the publ. intro., vss. 1–4 and 5–8 appear to be parallel hymns,
the first with multiple Dawns, the second with just one.
I.92.1: In pāda a my tr. differs slightly from the standard ones. I suggest that the
dawns have transformed themselves into a beacon, while in the standard interpr. they
have created/manufactured a beacon for themselves.
In b following Proferes 2003 (JAOS 123, pp. 330-31), I suggest, on the basis
of the fuller expression in the parallel verse 5c, that their “beam” is being anointed
like the sacrificial post (sváru-) at a ritual.
156
In c the poet exploits the syntactic ambiguity of the middle participle
niṣkṛṇvāná-. In the frame he takes it as a reflexive, “presenting themselves,” but in
the simile it is transitive “presenting arms.” The middle voice is still justified,
however, because the arms belong to the subjects of the participle.
I.92.3: The standard tr. are in agreement that árcanti only means “they sing” here and
that the meaning ‘shine’ for this verb stem is dubious. This judgment seems shortsighted and unresponsive to the poetics of this hymn, and even if ‘shine’ is not the
normal sense of árca-, the nominal derivatives like arcí(s)- ‘ray, flame’ (see arcís- in
5a) would allow a ‘shine’ sense under the proper circumstances. And these are the
proper circumstances. This verse-initial verb, following on two verses describing the
color, brightness, and beams of the dawns, would most naturally be interpreted with
a ‘shine’ sense. The simile that follows, involving women at work, would then cause
the audience to rethink this assumption, producing the interpretation ‘sing’, with the
women singing work songs at their labors. But ‘sing’ makes little sense for the frame
of the passage: do dawns ‘sing’? While ‘shine’ makes little sense for the simile: do
working women ‘shine’? As in 1c the poet uses the ambiguity of the verbal pivot to
craft two incompatible but verbally impeccable pictures, but this time he relies not on
syntactic ambiguity as in 1c but synonymity -- a pun on the verbal root underlying
the verb form. The pun may be even cleverer than so far presented. Although the
primary sense of árcanti for the dawns should be ‘shine’, it is possible that there is a
secondary sense ‘sing’, in that birds begin to sing at dawn and the ritual recitation
also begins at that time. Similarly, although ‘sing’ should be the primary sense for
the women in the simile, they may also be said to shine if they are glistening with
sweat from their labors. It should be noted that Ludwig thought the verb was
“doppelsinnig” (see Ge n. 3a).
I.92.4: bárjaha- in b is a hapax, but related to dual barjahyè in AV XI.8.14 in a list of
body parts, found between the ribs and the sides. Whitney/Lanman tr. ‘nipples’. The
consensus tr. ‘udder’ seems reasonable, esp. on the assumption that usrā́ ‘ruddy’
designates a cow. However, the purport of the simile is a little unclear. Since cows
don’t wear clothes, their udders are not covered to begin with and uncovering is
unnecessary. Ge suggests that the simile is proleptic: Dawn’s breast becomes as
visible as a cow’s full udder, but this interpretation requires that the verb
(‘uncovers’) that should mediate between frame and simile has been semantically
elided.
The simile in d is similarly “off,” since cows don’t have the capacity to open
their own pens. Macdonell (Hymns from the Rigveda, 1922) tr. “as when cows break
from their stalls,” which makes more sense, but again misrepresents the pivotal verb.
I.92.5: As noted before, vs. 5 seems to match vs. 1 and begin a new hymn parallel to
vss. 1-4, but adapted for a single dawn, not the multiples in vs. 1. The return of the
words bhānú- ‘beam’ and √añj ‘anoint’ (1b) is particularly salient, as also √śri (2d),
√arc (3a), and péśas- (4a).
157
The poet produces yet another complication of simile structure in cd. The
Daughter of Heaven in d is clearly the subject of the clause, but the simile occupying
c has a participle in the masc. nom. sg., añján, which cannot modify the grammatical
subject. We must here supply a masculine ritual officient, who is propping up and
anointing the sacrificial post, as Dawn props up and anoints her beam. (Note that
bhānúm served as object of both verbs in the earlier parts of the hymn: 1b bhānúm
añjate, 2d bhānúm … aśiśrayuḥ.) The two distinct objects in simile and frame are
mediated by the word péśas-, which is appropriate to both. Dawn has just donned her
own péśāṃsi in 4a, but péśas- can also be used of ritual paraphernalia -- e.g., the
pressing stones, which are called adhvarásya péśaḥ at VII.42.1.
I.92.6: vayúnā kṛṇoti reprises ákran … vayúnāni in 2c.
I.92.8: ‘Forelock’ for pravarga- may seem slightly flippant, but I wanted to capture
the ‘twist’ sense of prá √vṛj, in contrast to the anodyne ‘having slaves at the front’
found in the other tr.
The final bṛhántam of d obviously refers back to the rayím so carefully
detailed in ab, but is separated from that phrase by a considerable amount of material.
Rather than making it a long-delayed part of that NP I have taken it as a pseudo-goal
with vibhā́si, with vā́ja-prasūtā giving grounds for that goal.
I.92.8–9: vibhā́si (8c) and ví bhāti (9b) should have been rendered with the same
English verb, either “radiate” or “shine.” I would now change the publ. tr., to ‘radiate’
in both cases.
I.92.9: Unfortunately it is not possible to render the three fronted forms of víśva(pādas a, c, d) in non-awkward English.
cákṣuḥ in b can be either nom. or acc. Most tr. take it as the latter, with Dawn
either facing the eye(s) of, presumably, humans (so Ge [/WG], Re) or the eye of the
sun (so Macdonell [Hymns from the Rigveda], Maurer). The latter is unlikely on
pragmatic grounds: although the eye of the sun is a standard trope in the RV (see in
fact 11d), the sun rises after dawn, and she surely wouldn’t be looking behind her, to
the east. The former is certainly possible, but I prefer to take cákṣuḥ as nominative,
referring to Dawn as an eye, since she has just been said to oversee (abhicákṣyā) all
creatures.
I’m not entirely sure what pāda d refers to, but perhaps it means that by
dawning she has stimulated ritual speech and therefore “finds” it.
I.92.12: This last verse of the trimeter portion of the hymn plays more tricks with
simile structure. In pāda a prathānā́ functions like niṣkṛṇvānā́ in 1c; that is, it has
both reflexive and transitive values. In the frame it is reflexive (“spreading herself”),
in my view, though Ge, Re, Macd., Maurer all supply “rays” as the object, on no
contextual grounds. (However, Ge does allow for the reflexive possibility in his
notes 12a, 1c.) In the simile paśū́n is the object; the assumed subject might be
158
“cowherd” or the like (so tentatively Re). But the simile is quite appropriate to Dawn
for two reasons: first, it is often said that livestock go out to pasture at dawn and
return to the fold at night; second, the rays of Dawn are often referred to as cows (see
in fact vs. 2 above), and though I don’t think the stem paśú- is so used, it inhabits the
same semantic space as the more usual terms that participate in this metaphor.
The second simile, síndhur ná kṣódaḥ in b, is taken by Ge, Re, Macd., Maurer
with the verb of b, vy àśvait. Because this verb ‘whitened’ is intransitive, this
requires either taking kṣódaḥ adverbially (Re “comme le fleuve (en son) remous”) or
construing síndhuḥ … kṣódaḥ as a loose compound (Ge “wie die Stromflot”).
Although either is possible, the syntactic difficulties disappear if the simile falls
within the verbal domain of prathānā́, “like a river spreading its surge.” Given this
poet’s particular delight in manipulating similes, it is entirely in character to end this
section of the hymn with a triple play.
WG interpret the similes more or less as I do.
Pāda c is structurally parallel with 11c, using the same verb mī ‘diminish,
belittle’, though here in the negative (11c praminatī́, 12c áminatī), both playing off
āminānā́ in 10c.
I.92.14: “rich in … / richly…” implies that there is an etymological connection
between gomati, áśvāvati and revát, which of course there is not. But “possessing
cows and horses” seems too flat.
I.92.15: The hí + imperative followed by áthā + imperative is the same structure as
I.10.3; see the commentary thereon.
I.92.16: √yā is the usual verb that governs vártis- (I.34.4, I.116.8, etc. etc.), and I
supply a form of it here.
The publ. tr. implies that sámanasā is a vocative, but, by its accent, it’s a
nominative and would better be tr. “as like-minded ones.”
I.92.17: Gautama’s penchant for clever exploitation of syntactic possibilties returns
in this verse, after a series of unremarkable Uṣṇih verses (13–16). The verb in b,
cakráthuḥ, by my interpretation takes two different types of predicates and exhibits
slightly different root meanings: in pāda a “make your call to heaven,” that is, make
the call reach heaven; in b “make, i.e., create, light.” I am in agreement with the
standard tr. of pāda b, but diverge from them in a, where Ge (/WG) supply a separate
verb (“erhebt”) to govern ślókam and interpret ā́ diváḥ temporally (“zu dieser Stunde
des Tages”), not spatially. Although ā́ diváḥ is sometimes used temporally (e.g., in
the expression trír ā́ diváḥ), I take the ā́ here in the meaning “all the way to” (Gr’s
“Praep. mit folgendem Abl. … bis an, bis hin (räumlich)”). This usage of ā́ was
prepared for in the immediately preceding vs., 16a asmád ā́ “all the way to us”
(though in opposite order). For a closely parallel expression, in a Dawn hymn, cf.
III.61.4 ā́ntād [i.e., ā́ ántād] diváḥ ... ā́ pr̥thivyā́ḥ “all the way to the end of heaven
159
and of earth.” For the ślóka- reaching heaven, cf. I.83.6, 190.4, and for ślókam as
object of √kṛ, IV.53.3.
I.93 Agni and Soma
I.93.1: The expression in d bhávatam … máyaḥ echoes the compound in the final
verse of the immediately preceding hymn mahobhúvā (I.92.18) -- both dual, though
applying to different divine pairs, the Aśvins (I.92.18) and Agni and Soma (here). It
is possible that this final hymn of the Gautama sequence, with its unique dedication
to Agni and Soma (see publ. intro.), was attached just here because of the duplication
of phraseology. This connection would be clearer in the publ. tr. if they were tr. in
the same way. I would therefore change “become refreshment” to “become joy” in
this verse.
I.93.4: As noted in the publ. intro., Bṛsaya, with his very non-Indo-Aryan name, and
the destruction of his offspring are found elsewhere only in VI.61.3, a hymn devoted
to Sarasvatī. The theft of the provender (avasá-) from the niggard (paṇí-) is also
found in that hymn, in vs. 1. The connection between the two hymns to different
gods, found in different parts of the Saṃhitā, is not clear to me.
I.93.5: Pāda b contains an inverse vāyav indraś ca construction: agníś ca soma.
I.93.7: The three imperatives in b, vītáṃ háryatam … juṣéthām, all belong to roots
that ordinarily take accusatives, yet all must be construed with the fronted gen.
havíṣaḥ prásthitasya in pāda a. I don’t quite understand the case syntax, but assume
that these three verbs have been lumped together generically as verbs of consumption
and enjoyment, for which gen. complements are often possible.
In c bhūtám can be either imperative or injunctive and is taken as injunctive in
the published translation (as also by Ge and Re). However, since the immediately
preceding hymn contains the clausal diptych hí + imperative, followed by áthā +
imperative, exactly matching the structure here, I now think it would be better to
interpret bhūtám as impv. here as well: “Become providers …, then establish …” (So
WG.) An imperative interpretation also works better with the three abrupt
imperatives that precede in b.
I.93.8: saparyā́t in b echoes saparyáti in 2b, but displays the more usual case frame:
acc. (god) + instr. (means of service). The instance in 2b takes acc. (means of
service) + dat. (god), which is only otherwise found at X.37.1. The occurrence in this
verse thus functions as a species of poetic repair.
I.93.9: The purport of the final pāda isn’t clear to me, but it may indirectly comment
on the unusual nature of the joint dedication of the hymn to these two gods.
I.93.10: dā́śati echoes dā́śāt in 3b.
160
I.93.10–11: 2nd du. dīdayatam (10c) and jujoṣatam (11b) are somewhat problematic
forms. They belong to redupl. thematic stems; other forms to these stems are
subjunctives to the perfect. However, they have secondary endings, and the act. 2nd
and 3rd du. subjunctive endings should be primary. Lub identifies them as impvs. (Gr
also jujoṣatam); Kü (35) treats them as “hybrids” -- morphologically belonging to the
subjunctive but functioning as imperatives. They differ from some other secondarily
thematized pf. impvs. in having full grade of the root syllable, like the subjunctive
(versus zero-graade in pīpyatam √pi, √vāvrdhásva √vṛdh, etc. -- but cf. pipráyasva
√prī). In the publ. tr. they are treated as subjunctives (“you will shine,” “you will
enjoy”), but they would might be better rendered as imperatives.
I.94 Agni
I.94.1: BR corrected sám mahema to sám ahema ‘we would bind together’ to √nah,
an emendation accepted by Gr, Ge, Old (tentatively), KEWA I.153. Although I also
accepted it in Jamison 1983: 87, I now follow Gotō (1987: 243; accepted in EWA
s.v.) in seeing it as a form of √mah ‘bring about’, etymologically separate from
√mah ‘magnify’. Although I do not like multiplying entities, Gotō’s etymological
connections look reasonable; an interpretation with ‘make great’ is difficult; and a
thematic pres. or aor. stem aha- to √nah does not exist and the zero-grade (<*n̥h)
that would have to underlie is not otherwise found to √nah.
I.94.2: yásmai √yaj presents the classic parasmaipada configuration: “sacrifice for
someone else,” but āyájase is resolutely middle. I do not have a good explanation for
this beyond the fact that the functional distinction between act. yájati and mid. yájate
is not well developed in the RV, esp. with preverb ā́, and also that Agni, the subject,
is essentially sacrificing to himself.
Although act. sā́dhati can be transitive (and WG so tr. it), it also has
intransitive/absolute uses (cf. the same tag in VI.70.3). Moreover, in this hymn
sādháyati (3a, 4c) serves as the transitive counterpart; the poet seems to have
constructed an -ati /-áyati Grundverb/Kaus. relationship.
I.94.3: A choppy series of clauses, five in the first three pādas, with only the last
dependent in any way on another.
I.94.5: gopā́(ḥ) in pāda a can be either sg. or pl. All the standard tr. take it as the
former, referring to Agni, but this makes grammatical difficulties. Ge interprets the
following asya as a word-haplology for *asy *asyá, thereby getting the 2nd sg. copula
but gratuitously accenting the pronoun. Old reads *asy for asya and suggests
accenting *cáranti. Re takes viśā́ṃ gopā́(ḥ) as a 3rd sg. expression (”Agni est …”) and
seems to leave the following text alone. But all such tr. must introduce an accent on
whatever word they think follows the short initial clause, since, as it stands, two
unaccented words, asya caranti, would open that clause. These difficulties can be
161
avoided by interpreting gopā́ḥ as plural, subject of caranti and coreferential with
jantávaḥ. Pl. gopā́- is elsewhere used of Agni’s sidekicks, presumably his flames, in
VI.8.7. Though it is true that this same expression viśā́ṃ gopā́(ḥ) is found two hymns
later (I.96.4) as an undoubted singular referring to Agni, the grammatical
complications resulting from taking it as a singular here outweigh that argument for
me.
Most tr. take the two-footed and four-footed as appositional to jantávaḥ and
therefore as those who roam by night. This is contrary to the usual RVic vision of the
creatures settling down at night. Agni’s flames make more sense as the roamers,
evoking the flickering firelight that provides nightly illumination. The flames are
called “herdsmen” because people and animals gather around the fire then, like a
flock. This interpretation does raise the question of how to construe the nom. case of
that expression dvipác ca yád utá cátuṣpat. I take it as a definitional relative clause
expanding on viśā́m (so also Old). The ca … utá is a mix of two “both … and”
constructions: X ca Y ca and utá X utá Y, with the first half of the conjoined NP
fronted around the subordinating yád. This construction is very similar to 9b dūré vā
yé ánti vā ké cid, complete with the fronted first element.
The etymological relationship between citráḥ and praketáḥ in c is not easily
rendered in English.
I.94.10: All standard tr. take b vṛṣabhásyeva te rávaḥ as the main cl. to the yád cl.
beginning in a, while I take it as a part of that yád cl. There is no way to tell, since
the second cl. lacks a verb whose accent or lack thereof would settle the matter.
There are also no semantic implications; I weakly prefer mine because of the fairly
emphatic ā́d ‘then, just after that’ beginning the second hemistich.
All tr. take vanínaḥ as ‘trees’ (lit. ‘possessing wood’). This is doubtless the
primary sense here, since this sequence of vss. describes a forest fire, but ‘winners’ is
also possible, since the metaphorical context is that of a chariot race; see esp. 10a,
11c.
I.94.11: I supply ‘makes’ on the basis of 9c sugáṃ kṛdhi, though strictly speaking
“that is an easy way” is perfectly acceptable.
tāvaká-, the vṛddhi derivative of the genitive of the 2nd ps. pronoun, enlarged
with -ka-, is a striking form, though exactly what special effect it is aiming at is
unclear. For the use of -ka- in pronominal derivatives, see Jamison 2009 (IIJ 52). As
discussed there, the -ka- tends to signal a lower register or slangy tone.
I.94.12: Most tr. take Mitra and Varuṇa as the objects of dhā́yase: approx. “Agni is
for the suckling of Mitra and Varuṇa.” But Agni is the one more likely in need of
suckling (that is, feeding the fire); cf., e.g., II.5.7. In particular, in the next hymn,
I.95.1, Agni is the object of the transitive/causative dhāpayete. I therefore take the
gen. mitrásya váruṇasya as dependent on héḷaḥ in b, parallel to marútāṃ. héḷa(s)- is
characteristic of Varuṇa (c. I.24.14, VII.84.2; in VII.62.4 Mitra is included).
162
avayātā́m is the problem in the second pāda. Most tr. take it as a 3rd sg.
imperative (e.g., Re “qu’il exorcise”), but if so, it must be a middle root pres.
imperative, and there are no middle forms to this root √yā (as opposed to √yā
‘implore’). There is also the problem that the form should not be accented on the
stem but on the preverb (*áva yātām; see Ge n. 12b). Lub. appears to take it as a gen.
pl. of the act. part. to the root pres., but in the absence of a tr. it is hard to know how
he would interpret it in context. Best to follow Old (who follows BR) in emending to
*avayātā́, the nom. sg. agent noun. Re is tempted by this reading, save for the fact
that with the acc. complement héḷaḥ we should expect the accent ávayātā. However,
the “rule” about the case of complements to agent nouns is broken so often that this
objection is not cogent. The final -m could have been acquired from the following
marútām (so Old), and since √yā ‘drive’ is a characteristic Marut verb and a genuine
example of the gen. pl. part. yātā́m is used of the Maruts in the refrain V.55.1–9, it
may have been rhetorically natural to transform the agent noun into this participle.
I.94.13: Both a and b turn on interpreting one word as both an appellative and a
proper name: mitráḥ in a, vásuḥ in b, though they occupy different places in the
parallel structures.
I.94.15–16: The meter here changes to Triṣṭubh, and the intra-hymnic refrain is
dropped. The 2nd hemistich of vs. 16 contains the first example in the Saṃhitā of the
Kutsa refrain.
I.94.15: Contra standard tr., I do not take anāgástvam as the obj. of dádāśaḥ, since
√dāś seldom takes an object that is not a god, and when it does, it is a ritual offering.
Instead I take it as belonging to a parallel nominal clause, still in the domain of
yásmai.
The final phrase that is the ostensible main clause, té syāma “might we be
they,” violates number agreement: pl. té cannot properly pick up sg. yásmai. This is a
variant of a common tag and feels tacked on.
I.94.16: Because sá with 2nd ps. reference is restricted by rule to imperatives (see
Jamison 1992 [HS 105]), ab must be a single clause, pace Ge.
I.95 Agni
I.95.1: Ge suggests that the sun is referred to in c, Agni in d. But it seems more likely
that Agni is the referent of both (or at least the first referent: since Agni = Sun is a
common identification, there may be secondary association), and the point is simply
that fire looks different in the daylight from at night.
I.95.3: The riddles in this verse have given rise to much more speculation than can be
covered here. In the first hemistich the major question is whether the three births of
pāda a (trī́ṇi jā́nā) are matched by three birth locations in b or only two. That there
163
are three locatives (samudré … diví … apsú) might support the first view; that there
are only two ékam’s the second. Despite their polarized positions in the pāda, I
believe that samudré … apsú refer to one place of birth, diví to another. X.45.3,
where samudré … apsú is one place of kindling (pāda a), diváḥ … ū́dhan a second
(b), and a third is referred to in cd, supports the two-locale view. This then allows the
third birth to be the ritual kindling described in the first vss. of the hymn, while very
tentatively we might identify the birth in waters as that of Apāṃ Napāt and that in
heaven as that of the sun. The three births can also refer to the three ritual fires, and
that notion is taken up obliquely in the second half of the verse, in my opinion.
The second hemistich describes the carrying of the kindled fire to the east end
of the ritual ground, to become the Āhavanīya fire (not so called in the RV, but
clearly referred to often in the text). This pacing out of the ritual ground also
establishes the other ordered elements in the ritual, hence the VPs of pāda d.
I.95.4: As indicated in the publ. intro., this vs. describes the kindling of the fire in
typically enigmatic fashion, with multiple referents possible for each entity. (See the
various tr. and comm. for disc.) It also contains in pāda b a version of the beloved
paradox in which the child gives birth to his parents, though in this particular case I
am uncertain what exactly is meant. The janayata in 4b picks up the janayanta in 2a,
but with the subject and object reversed: in 2a the ten (fingers) begat the infant (fire);
here the calf (fire) begets his mothers. On the basis of the echoing of 2a, I assume
that the mothers here are the fingers, though waters are also a popular suggestion -one that does not fit the ritual context, in my opinion. The hard-working women in c
are probably also the fingers, though kindling sticks are also possible, esp. on the
basis of 5b.
svadhā́(-van)- is a signature word of this part of the hymn: 1c, 4b, 4d.
I.95.5: The hymn began with unnamed dual feminines attending to the young fire;
this verse also contains unnamed dual feminines in the same role. But the identities
of the two pairs are different: Night and Dawn in 1, the world-halves in 5cd, 6a.
Note the word play of pratīcī́ … práti, with slightly different meanings.
I.95.5–6: The caus. (práti) joṣayete in 5d, 6a literally means “they two cause (him) to
take pleasure (in themselves),” but this tr. seems too heavy, esp. in adjacent pādas.
I.95.6: The position of simile-marking ná is aberrant, preceding the simile’s noun
méne and following the verb. Since the full simile is bhadré … ná méne, it seems that
this syntagm was distracted in some fashion, with the verb placed in its middle. On
méne see I.62.7.
The etymological figure in c, dákṣāṇāṃ dákṣapatiḥ “skill-lord of skills,” then
participates in a phonological figure with semantically distinct dakṣiṇatáḥ “from the
right (/south)” in d.
164
I.95.7: I take the first half-verse as a description of fire both sending its flames
upwards and stretching out horizontally. Most tr. interpret yatate in b as transitive
(e.g., Ge “er eifert beide Heeresreihen an”), but the medial forms of this pres. stem
are intransitive or reflexive. I instead take ubhé sícau as an acc. of extent and assume
that it refers to the seams between the two world-halves found in vss. 5–6. With these
two halves, heaven and earth, meeting at the horizon, as it were, the seams between
them would stretch horizontally.
The “new clothes” he leaves for his mothers in d are probably the ashes that
fall on the kindling wood as he burns it.
I.95.8: The first three pādas of this verse sounds very somian: the cows [=milk] and
the waters of b are the standard ritual substances mixed with soma in the IXth
Maṇḍala, the beginning of a, tveṣáṃ rūpáṃ kṛṇute, is also found in a soma hymn,
IX.71.8, and √mṛj ‘wipe, groom’ in c is a signature soma verb. This may be an
example of the covert equation of the two ritual gods, Agni and Soma. Pāda a is
easily interpreted in an Agni context; b makes more difficulties: the cows can here be
the ghee poured into the fire (which could cause the creation of an “even higher”
form in a), but the waters would obviously have a diminishing effect on the fire if it
actually “mingles” with them. Ge (n. 8b) suggests that the vedi is being sprinkled
with water, but that’s not what the text says. I think better to assume that this is an
incompletely adapted somian description.
Pāda c has caused interpretational difficulties, in particular because of the
apparent equation of the poet (kavíḥ) and the insight (dhī́ḥ), both nominative and both
apparent subjects of marmṛjyate. Ge takes them as conjoined nouns with suppressed
conjunction, “the poet (and) his thought,” which are both tending to the Unterlage of
Agni. This is not a bad solution, but it assumes that the kaví- is the human poet. Since
Agni was identified as a “great poet” (mahā́n kavíḥ) in 4d, it’s desirable to maintain
this identification if at all possible. Another solution is to take dhī́ḥ as a separate
nominal clause, as Re does (“il est la Vision-poétique”). He takes the referent of both
kavíḥ and dhī́ḥ to be Agni. WG seem to take dhī́ḥ as the object of the verb, though
without comment. This is presumably inspired by the fact that acc. pl. dhiyáḥ does
elsewhere serve as object to marmṛjyáte (IX.47.4). However, making dhī́ḥ into an acc.
pl. is grammatically difficult (though, I suppose, not impossible).
My solution involves reading the intensive marmṛjyate simultaneously in two
syntactically different ways, both of which are paralleled elsewhere in the RV. The
stem marmṛjyáte is one of the new-style -yáte intensives, which have medial
inflection and passive accent even if they have active semantics and pattern with
active stems in their averbo. See, for example, IX.47.4 just cited. (On this type and
its origin, see Jamison 1983 [MSS 42: 41–73].) Thus, the poet can be stroking the
budhnám in this transitve interpretation, like the many active forms of this root
including the act. athem. intensive mármṛj-. But several instances of medial
marmṛjyá- have the passive value the morphology implies (e.g., IX.62.13
marmṛjyámāna āyúbhiḥ, of soma), and I take dhī́ḥ as the subject of the verb read this
time as a passive. This syntactic pun might be clearer in the publ. tr. if the same word
165
had been used for both readings; better might be “… keeps stroking … is being
stroked,” though ‘groom’ is actually better with the insight as subject.
I.95.9: Ge (/WG) assume that pāda a treats the Paryagnikaraṇa or the ritual act of
circling around an object with a firebrand, but jráyaḥ ‘expanse, expansion’ seems to
me rather to depict the horizontal spread of the fire out from its original kindling
place.
I.95.10: In a srótaḥ can be either nominative, as appositive to the underlying subject
Agni, or accusative, an object parallel to gātúm ūrmím. I chose the second alternative,
contra Ge / Re, but either is possible and the meaning is virtually identical.
I.95.11: revát in b may be either adverbial (as I’ve taken it) or the object, with
supplied noun, of ví bhāhi (so, e.g., Ge “Nun strahle … reiches (Gut) aus”). Again
either is possible and the effect is essentially the same.
I.96 Agni
The first verses of this hymn connects Agni with the semi-divine ancestors
and culture heroes of the Ārya: Āyu, Manu, Vivasvant (vs. 2), Bharata (vs. 3),
Mātariśvan (vs. 4).
I.96.1: Most tr. interpret mitrám … sādhan as “they conclude an alliance” (so Re; Ge
“Freundschaft”), but since mitrá- is so often a descriptor of Agni as the go-between
between gods and men, I assume that Agni is the referent here as well. So also Old
SBE.
I.96.2: This verse attributes primal creative power to Agni first in the ritual realm
(ab), then in the cosmic realm (c). Ge (/WG) supply a different verb in c,
“bescheint”), presumably because Agni is not usually credited with cosmogonic
powers. But there is no contextual support for a new verb here, and in parallel
clauses containing only one verb the default interpretation is to supply the same verb
with the second clause. Given how much generative power is ascribed to the Vedic
sacrifice, it is not surprising that Agni’s ability to engender Manu’s people, that is,
those who sacrifice like Manu, can then be extended to his ability to beget the major
cosmic features -- esp. as in this enterprise he is partially identified with the sun.
Note that in 4c he is explicitly named as “the begetter of the two worlds” (janitā́
ródasyoḥ) as well as “the finder of the sun” (svarvít).
I.96.3: On the Bharata fire as “the focus of worship of multiple ārya clans,” see
Proferes 2007:37.
Though sṛprá- receives various tr., it must be related to sarpís- ‘melted
butter’. See EWA s.v. sarpíṣ-; it has nothing to do with √sṛp ‘creep’.
166
I.96.4: Contrary to the standard tr., I take pāda a as containing a separate predication:
in his capacity as Mātariśvan he ensures prosperity, while as Sun-finder he finds the
way.
I.96.7: The full integration of the refrain into the syntactic structure of the verse is
signalled by the shift from nominative reference to Agni to accusative, clear first in
kṣā́m in b, since sádanam in a is neuter and could be in either case -- the common use
of a morphologically ambiguous form as pivot.
I.96.8: With the refrain having finally been integrated into the verse in 7, it is
abruptly dropped and its final and most salient word, draviṇodā́- explodes in vs. 8.
Though sánara- is a hapax, ‘having superior men’ seems a fairly safe bet, esp.
given semantically parallel vīrávant- in the next pāda.
I.97 Agni
The grammatically incomplete refrain of this hymn is introduced first as the
first pāda of this hymn and then repeated as 1c. In that verse, the refrain’s participial
construction is integrated into the verse structure, modifying the subject of the main
verb śuśugdhi in b. This integration is not found again until vs. 6 and continues
thereafter till the end of the hymn (vss. 7–8). However, the semantic distance
between the verse proper and the refrain narrows as the hymn progresses. In vss. 2-3
there is no explicit mention of Agni, but in vss. 4–5 he appears, as enclitic pronoun +
voc. (te agne) in 4 and gen. (agnéḥ) in 5, though neither is in the appropriate case to
match the refrain.
I.97.1: It does not seem worthwhile to try to replicate the difference between the two
reduplicated forms, the intensive participle (śośucat) and the perfect imperative
(śuśugdhi) in tr. Re’s tr. of the inten. part., “écartant-par-ton-éclat-puissant,” is a
cautionary example of why. WG’s “immer wieder wegflammend” is somewhat less
clunky but still gets wearisome in repetition.
I.97.3–5: These three middle verses, before the refrain becomes reintegrated into the
verse, begin identically: prá yát, though the sense of yád in 3–4 differs from that in 5.
I.97.3: The referent of bhándiṣṭha eṣām “the most fortunate one of these” is not clear.
However, since he is conjoined with “our patrons” and patrons are often conjoined
with “us” (as in the next verse), referring to the poet-performers (explicitly II.2.12
stotā́raḥ … sūráyaś ca), it is likely a singer or poet, perhaps even this very poet,
referring to himself in the 3rd ps.
I.98 Agni
I.98.2: As noted in the publ. intro., I take this verse as an allusion to the well-known
story of Agni’s disappearance and the gods’ search for him (treated in detail in X.51–
167
53). (So also Old SBE.) However, this mythic allusion must be mediated by
reference to the here-and-now, given the hope for Agni’s protection expressed in the
final pāda. This mediation is perhaps signaled by pāda b, where Agni’s hiding place
is identified. Instead of the waters, where Agni takes refuge in the myth, he has
entered “all plants” (víśvā óṣadhīḥ). Agni’s hidden presence in plants, the quality that
allows him to be born from them, is a standard trope in hymns treating the kindling
of the ritual fire and is in fact alluded to at the end of a nearby Kutsa hymn, I.95.10d
antár návāsu carati prasū́ṣu “he roams within the new, fruitful (plants).”
I.99 Agni
The only single-verse hymn in the RV, it closes Kutsa’s Agni cycle. Though
attributed to Kaśyapa Mārīca by the Anukramaṇī, it shows clear connections with
other parts of Kutsa’s Agni hymns, for which see publ. intro.
168